Research seminars

Seminario: "Un informático trabajando entre biólogos en el CNAG (Centro Nacional de Análisis Genómico). "

Fecha: 22/05/2017 Hora: 12:00

Lugar: Aula 1 del Ada Byron

Ponente: Santiago Marco-Sola

RESUMEN:

En biotecnología, el reciente desarrollo de la secuenciación de alto rendimiento o HTC (high-throughput sequencing) ha supuesto uno de los avances más relevantes para la investigación moderna en biología y biomedicina. Este conjunto de protocolos y tecnologías permiten la secuenciación del genoma de un individuo en cuestión de días. Sus aplicaciones van desde la medicina diagnóstica pasando por la biología evolutiva hasta la investigación bioquímica y molecular. De este modo, la secuenciación se ha convertido en el caballo de batalla de multitud de grupos de investigación, protagonizando muchos de los descubrimientos científicos más relevantes de la última década.

Actualmente, el CNAG (Centro Nacional de Análisis Genómico) es capaz de secuenciar hasta 1.5 terabytes de información genómica al día. Equipados con 3472 nodos de computo y 7.6 petabytes de almacenamiento, el CNAG se encarga de procesar, analizar, clasificar y almacenar billones de secuencias de ADN diariamente. Para hacer frente a este reto, muchos grupos de investigación en algoritmia y computación investigan nuevos algoritmos y metodologías de análisis, con el objetivo de desarrollar herramientas eficientes, escalables y flexibles para el análisis genómico en bioinformática.

En esta charla presentaré un resumen de los protocolos, herramientas y algoritmos bioinformáticos que un centro de secuenciación masiva como el CNAG emplea rutinariamente. Del mismo modo, introduciré los principales retos computacionales a los que nos enfrentamos, futuras tendencias en bioinformática y biotecnología, así como nuevas e interesantes oportunidades para informáticos interesados en formar parte del mundo de la bioinformática.

PONENTE:

Santiago Marco-Sola se licenció en ingeniería informática por la Universidad de Zaragoza (Centro Politécnico Superior) en el 2010. Posteriormente, obtuvo el máster en computación por la UPC (Universidad Politécnica de Barcelona) en la especialidad de algoritmia y programación. Durante los últimos 6 años, ha trabajado en el "Bioinformatics Development & Statistical Genomics Team" del CNAG (Centro Nacional de Análisis Genómico). Durante este periodo desarrolló su tesis doctoral centrada en técnicas eficientes de alineamiento de secuencias genómicas aplicadas a la secuenciación de alto rendimiento (high-throughput sequencing).

Actualmente es profesor en la Escola d'Enginyeria de la UAB (Universidad Autónoma de Barcelona) e investigador en el grupo CAOS (Computer Architecture and Operating Systems). A su vez colabora con el grupo "Computational Biology of RNA Processing" en el CRG (Center for Genomic Regulation) en el desarrollo de herramientas de alto rendimiento para análisis de datos genómicos en bioinformática. Sus intereses de investigación incluyen algoritmos de indexación y alineamiento de secuencias, algoritmos de compresión y aplicaciones de alto rendimiento en entornos heterogéneos.

Seminario: " Ajuste automático de sistemas, algoritmos y simulaciones mediante optimización Bayesiana"

Ponente: Rubén Martínez Cantín (Centro Universitario de la Defensa, SigOpt)

Lugar: Seminario A-25. Edificio Ada Byron

Hora: 11:00 - 18 mayo

Abstract:

La optimización Bayesiana combina aprendizaje automático y técnicas de decisión óptima para proporcionar un método de optimización altamente eficiente. Permite obtener resultados similares a los de otros métodos de optimización global (algoritmos genéticos, simulated annealing, swarm intelligence...) con sólo una mínima fracción de las muestras, lo que permite ser utilizado en procesos muy costosos. Recientemente, su uso está creciendo de manera exponencial, tanto a nivel académico como industrial, gracias a las numerosas aplicaciones que se están encontrando: desde el ajuste óptimo de parámetros de algoritmos, control de sistemas y robots, diseño de simulaciones, mejora de procesos de fabricación, diseño industrial, etc. Especialmente, el ajuste automático de modelos de "deep learning" está teniendo gran repercusión. En la charla se dará una revisión de la metodología, mostrando los avances recientes realizados dentro del grupo de investigación.

Bio:

Rubén es profesor del Centro Universitario de la Defensa y miembro del grupo de RoPeRT del I3A. Su investigación en los últimos años se ha centrado en el desarrollo de optimización Bayesiana y su aplicación al aprendizaje automático y la robótica. Actualmente es investigador en SigOpt Inc., que ofrece BoaaS: "Bayesian optimization as a Service".

Charla:Programming Molecules in the Age of Nanotechnology

Tendrá lugar el lunes 8 de mayo a las 12:00 en el aula A13 del Ada Byron

Abstract

When scientists combine computer science with the information-processing power of molecules, science fiction becomes a reality. Self-assembling, programmable systems at the nanoscale are poised to have a major impact on society, from personalized medical therapeutics to biosensors that could detect pollutants in our water or disease in your body. This talk will describe our work at Iowa State University aimed at using computer science and software engineering methods to design molecular programmed systems that are efficient, verifiably correct, and safe for use.

Short bio

Dr. Robyn Lutz is a professor of computer science at Iowa State University. She was on the technical staff of Jet Propulsion Laboratory, California Institute of Technology until 2012, most recently in the Software System Engineering group. Her research interests include safety-critical software systems, product lines, and the specification and verification of DNA nanosystems. She is an ACM Distinguished Scientist. She served as program chair of the 2014 International Requirements Engineering Conference, recently completed her second term as an associate editor of IEEE Transactions on Software Engineering, and is on the editorial board of the Requirements Engineering Journal.

Seminario: MIPSfpga: Using a Commercial MIPS Soft-Core in Computer Architecture Education

Tendrá lugar este viernes 5 de mayo, de 10:00 a 11:00 en el Aula A.05 del Edif. Ada Byron

SUMMARY: In this talk I will introduce MIPSfpga and its accompanying set of learning materials. MIPSfpga is a teaching infrastructure that offers access to the non-obfuscated RTL source code of the MIPS microAptiv UP processor. The core is made available by Imagination Technologies for academic use and is targeted to an FPGA, making it ideal for both the classroom and research. The supporting materials and labs focus on hands-on learning that emphasizes computer architecure, System on Chip (SoC) design and hardware-software codesign. Among other things, students learn to set up the MIPS soft-core processor on a field-programmable gate array (FPGA), run and debug programs on the core in simulation and in hardware, add new peripherals to the system, understand the microarchitecture and extend it to support new features, experiment with different cache sizes and content management policies, add new instructions using the CorExtend interface available in MIPS processors, and understand SoCs in embedded systems and how they are designed and built up in layers to run complex software such as Linux.

SHORT BIO: Daniel Chaver received the degree in physics from the University of Santiago de Compostela, Spain, in 1998, and the Electrical Engineering degree and Ph.D. from the Complutense University of Madrid, Spain, in 2000 and 2006, where he is currently an Associate Professor. He has taught many courses related to Computer Science and Electrical Engineering since 2000. He has co-advised 3 PhD thesis and has co-authored more than 40 papers. Since 2015 he collaborates with Imagination Technologies. His current research interests include: (1) architectural techniques for managing efficiently the memory hierarchy and (2) OS scheduling techniques for asymmetric multiprocessors.

Seminario: PMCTrack: delivering hardware monitoring support to the system software

Tendra lugar este jueves 4 de mayo, de 12:00 a 13:00 en el Aula A.05 del Edif. Ada Byron

Summary:

Hardware performance monitoring counters (PMCs) have proven effective in characterizing application performance. A large body of work has demonstrated that several components of the operating system (OS), such as the scheduler, can perform effective runtime optimizations in multicore systems by leveraging performance-counter data. While existing tools greatly simplify the collection of PMC data from user space, they do not provide an architecture-agnostic mechanism that is capable of exposing high-level PMC metrics to the OS. Thus, the implementation of OS-level PMC-driven optimization schemes is typically tied to specific processor models.

In this talk I will present PMCTrack, an open-source tool for the Linux kernel that seamlessly enables the system software to access PMC data in an architecture-independent fashion, and also provides other insightful monitoring information available in modern processors, such as cache occupancy or energy consumption. Despite being an OS-oriented tool, PMCTrack still allows the gathering of monitoring data from user space, making it possible for users and developers to perform offline analysis in various ways.

Short bio

Juan Carlos Saez received his Ph.D. in computer science in 2011 from the Complutense University of Madrid (UCM), where he also obtained the Extraordinary Doctorate Award. He is now an Associate Professor in the Department of Computer Architecture at UCM. Since 2013, he has served as the UCM Campus Representative of USENIX, the Advanced Computing Systems international association. In the last few years, he has been teaching different courses related to Operating Systems and Computer Architecture. His research interests include energy-aware computing and improving the interaction between the system software and hardware for emerging architectures. His recent research activities focus on OS scheduling on heterogeneous multicore processors, exploring new techniques to deliver better performance per watt, and quality of service on these systems.

Seminario:Real-time Augmented Reality with Deformable Objects: Systems and Applications

Lugar: Seminario 25 Ada Byron.

Hora: 11:15 27 abril 2017..

Ponente: Toby Collins

Abstract.

An important yet unsolved problem in computer vision and Augmented Reality (AR) is to register the 3D shape of deforming objects with live 2D videos. This has important applications in Augmented Reality (AR), Computer-Assisted Intervention (CAI), Computer Graphics and 3D scene understanding. At the CNRS Endoscopic and Computer Vision (EnCoV) lab I worked on this topic with Prof. Adrien Bartoli to develop robust, real-time systems. In this talk I will discuss two state-of-the-art systems. The first is Laparaug, a system to improve laparoscopic surgical intervention in gynecological surgery. This is designed to help the surgeon locate hidden sub-surface structures, including vessels and tumours. The structures are first segmented in a pre-operative MR or CT scan, and then augmented onto the laparoscope's live video in real-time to guide the surgeon. The second system is a state-of-the-art approach to densely track generic deforming objects in 2D videos, with both medical and non-medical applications. This was the first dense approach to be demonstrated live, at the International Symposium of Mixed and Augmented Reality (ISMAR) 2015 and The European Conference on Computer Vision (ECCV) 2016. I will also discuss the main open objectives and future trends in the field.

Seminario:Image Guided Surgery: The visible patient in the Hospital of the future.

Lugar: Seminario 25 Ada Byron.

Hora: 10:00

Día: 27 abril 2017.

Ponente: A. Hostettler



Seminario: "4D Vision from Images: Data-based vs. Model-based Learning"

Este viernes 10, de 17:00 a 19:00 en el seminario A.25 Ed. Ada Byron tendrá lugar el seminario que se detalla a continuación. La actividad está enmarcada dentro del programa de invitación de profesores externos.

4D Vision from Images: Data-based vs. Model-based Learning

Antonio Agudo, Investigador Postdoctoral

Instituto de Robotica e Informatica Industrial (IRI)(CSIC)

Abstract #1

This paper addresses the problem of 3D human pose estimation from a single image. We follow a standard two-step pipeline by first detecting the 2D position of the N body joints, and then using these observations to infer 3D pose. For the first step, we use a recent CNN-based detector. For the second step, most existing approaches perform 2N-to- 3N regression of the Cartesian joint coordinates. We show that more precise pose estimates can be obtained by representing both the 2D and 3D human poses using N ⇥N distance matrices, and formulating the problem as a 2D-to-3D distance matrix regression. For learning such a regressor we leverage on simple Neural Network architectures, which by construction, enforce positivity and symmetry of the predicted matrices. The approach has also the advantage to naturally handle missing observations and allowing to hypothesize the position of non-observed joints. Quantitative results on Humaneva and Human3.6M datasets demonstrate consistent performance gains over state-of-the-art. Qualitative evaluation on the images in-the-wild of the LSP dataset, using the regressor learned on Human3.6M, reveals very promising generalization results.

Abstract #2

We present an approach to reconstruct the 3D shape of multiple deforming objects from incomplete 2D trajectories acquired by a single camera. Additionally, we simultaneously provide spatial segmentation (i.e., we identify each of the objects in every frame) and temporal clustering (i.e., we split the sequence into primitive actions). This advances existing work, which only tackled the problem for one single object and non-occluded tracks. In order to handle several objects at a time from partial observations, we model point trajectories as a union of spatial and temporal subspaces, and optimize the parameters of both modalities, the non-observed point tracks and the 3D shape via augmented Lagrange multipliers. The algorithm is fully unsupervised and results in a convex formulation which does not require initialization. We thoroughly validate the method on challenging scenarios with several human subjects performing different activities which involve complex motions and close interaction. We show our approach achieves state-of-the-art 3D reconstruction results, while it also provides spatial and temporal segmentation.

Seminario: NICP: Dense Normal Based Point Cloud Registration

Lugar: Salón de Actos, Edificio Ada Byron

Hora: 9:15

Ponente: Giorgio Grisetti

In this talk we present a novel on-line method to recursively align point clouds. By considering each point together with the local features of the surface (normal and curvature), our method takes advantage of the 3D structure around the points for the determination of the data association between two clouds. The algorithm relies on a least squares formulation of the alignment problem, that minimizes an error metric depending on these surface characteristics. We named the approach Normal Iterative Closest Point (NICP in short). Extensive experiments on publicly available benchmark data show that NICP outperforms other state-of-the-art approaches.

Seminario: Vision-based Perception for (Aerial) Robots

Lugar: Salón de Actos, Edificio Ada Byron

Hora: 10.00

Ponente: Margarita Chli, ETH, Zurich

Robots In a bid to equip robots with increased autonomy, this talk will motivate and describe our current research and future directions at the Vision for Robotics Lab (http://www.v4rl.ethz.ch), all the way from single-robot navigation to collaborative perception in a robotic team.

[Diis] Seminario de investigación: "Identification of timed Discrete Event Systems: an approach based on Time Petri nets"

Título: Identification of timed Discrete Event Systems: an approach based on Time Petri nets

Ponente: Francesco Basile, University of Salerno, Italy

Día: Jueves, 19 de enero de 2017

Hora: 09:30

Lugar: Seminario del Departamento de Informática e Ingeniería de Sistemas

Duración: 1 hora

==============

Abstract: The data collected from the observation of a discrete-event system are usually given in terms of behavioral sequences that may be fixed or may be increased in the course of the system operation (e.g., due to new experiments or simply to the system running). These data can be used to obtain a nominal model of the system, or, if it is already available, it may reveal not consistent with these additional observations if they include anomalies. This seminar presents an approach to system identification and repair of the system model based on time Petri nets

Francesco Basile received the Laurea degree in Electronic Engineering in 1995 and the Ph.D. degree in Electronic and Computer Engineering in 1999. From May 1996 to February 1997, he worked as Process Engineer at the Proma Group, a multinational company specialized in metal working for automotive industry. In 1999 he was visiting researcher for six months at the Departamento de Ingenieria Informatica y Systems of the University of Saragoza, Spain. From January 2000 to December 2001 he has been Post-Doctoral Fellow at the University of Naples. From 2002 to October 2014 he has been assistant professor of Automatic Control at the Dipartimento di Ingegneria dell'Informazione e Ingegneria Elettrica of the University of Salerno, Italy. He is Associate Professor of Automatic Control at the Dipartimento di Ingegneria dell'Informazione e Ingegneria Elettrica of the University of Salerno. He has published more than 100 papers on international journals and conferences. His current research interests are: modeling and control of discrete event systems, industrial automation and robotic. He has been member of the scientific committee of many international conferences.

He is General Chair of 14th International Workshop on Discrete Event Systems (WODES 2018). He has been member of the editorial board of International Journal of Robotics and Automation and IEEE Transactions on Control Systems Technology up to December 2016. He is a member of the editorial board of IEEE Transactions on Automation Science and Engineering, IEEE Control Letters since December 2016 and of IEEE Control System Society Conference Editorial Board since September 2007.

Seminario de investigación: Sequence Planner– integrating research results in practice

Título: Sequence Planner– integrating research results in practice

Ponente: Kristofer Bengtsson, Chalmers University of Technology, Gothenburg, Sweden

Día: Martes, 17 de enero de 2017

Hora: 11:00

Lugar: Seminario del Departamento de Informática e Ingeniería de Sistemas

Duración: 1 hora

==============

Resumen: It is a big challenge for us, researchers in the discrete event systems community, to use our research results in practice. There are a number of reasons for this, for example that it is hard to model the problems (and usually impossible to do for people not in the research group), and that it is hard to solve real world problems. But another reason is that research result usually requires significant effort from the practitioners as well changes to their working procedure, which makes it hard for us researcher to motivate them to start using our tools.

The last 8 years, our research group has been working with a software tool called sequence planner. The aim was initially to improve the modeling and analysis of automation system including verification, synthesis and optimization. But the tool was not used by practitioners until we included simple services for aggregating data from robotized automation system as well as emergency departments at hospitals. We developed a simple yet powerful concept called the tweeting factory (which also works really well in e.g. healthcare systems) for visualizing and analyzing live data. The interesting effect of this was that the use for our latest research results suddenly was of interest to the users.

During the talk, I will present the tool and some of the results we have had the last years. Especially the live data visualization, prediction services as well as energy optimization of robot stations where we can save up to 30% of the energy consumption of a robot. ==============

Kristofer Bengtsson received his Ph.D. degree in signals and systems from Chalmers University of Technology, Gothenburg, Sweden, in 2012. From 2001-2005, he was with Advanced Flow Control AB developing control systems and user interfaces, and from 2005 to 2011 with Teamster AB, an automation firm in Gothenburg. Currently, he is with the automation group in the systems and control division at Chalmers as well with Sekvensa AB, a research consulting firm. His current research interest includes design methods and computer solutions for logic control development and algorithms and methods for control of automation and healthcare systems: like sequence planning, virtual preparation and commissioning, distributed control, reinforcement and machine learning, error recovery and restart, analysis and visualization as well as robot energy minimization.

Seminario: Software Platform for Planning and Control of multi-robot systems by using High-Level Specifications

Ponente: Emanuele Alberto Vitolo, Master student at University of Salerno, Italy

Día: Miércoles, 21 de diciembre

Hora: 12:00

Lugar: Seminario del Departamento de Informática e Ingeniería de Sistemas

Duración: 1 hora

Resumen: Planning and controlling mobile robots is a research area that continues to receive a lot of attention. An intensively studied problem in this area is the navigation of a mobile robot, where a control strategy should be automatically generated such that the robot reaches a target position without colliding with obstacles. Some solutions proposed for this problem directly work in a continuous state space (e.g., navigation and potential functions), while others abstract the initial problem to a discrete finite-state one and find a solution by a search on a graph (e.g., cell decomposition methods). The latter methods were successfully tailored such that more complicated problems can be solved, where the specification is close to natural language and it is given as a formula combining temporal and logic operators. The propose of this work is to develop a multi-robot platform using Arduino robots for automatically computing trajectories for robots in order to fulfill high-level specifications and control the robots in order to follow the computed trajectories. In this first version of the platform, only classical navigation problem has been implemented (reach some final targets by avoiding the obstacles) but will be extended to more complicated missions. The main steps of the solution consist in:

(1) Discretization of the environment by using a cell decomposition approach. In particular Squared Cell Decomposition method has been implemented dividing the working space into square regions of a certain size which then will be set as walkable or not (depending by the presence of obstacles in that region).

(2) Use a discrete event system obtained from the previous decomposition and apply the A* algorithm to calculate paths for each robots to reach final targets.

(3) Establish a communication protocol between a computer and the Arduino mobile robots in order to update dynamically the actual positions and orientations in the environment. These information are provided by a camera placed on top of the workspace, which detects the robots and provides their position in real-time. To implement this communication XBee antennas have been used (which transmits data through a WiFi signal) and a communication protocol has been implemented to recognize and read the individual packets sent from the computer.

(4) Develop internal controllers for the Arduino robot, to be able to move within the workspace following the path and at the same time avoiding the obstacles.

(5) Finally, use the Petri Net models to implement a collision avoidance mechanism. For collision avoidance, waiting modes for the robots can be implemented in order to ensure maximum one robot in each region. However, this strategy may yield to deadlocks and a structural approach based on a Petri net system is used to develop a supervisory control to avoid these situations.

The platform has been completely implemented in c++ using an oriented object programming. Some videos illustrating real experiments of the developed platform will be provided.

Seminario: Despliegue de Servicios en Redes Vehiculares

Ponente: Miguel Báguena

Día: martes 20 de diciembre

Hora: 12:00

Lugar: Laboratorio de Telemática (EUPT, Teruel)

Duración: 2 horas

Abstract

Las Redes Vehiculares se van a convertir en una realidad en nuestras carreteras en un futuro cada vez más cercano. No obstante, todavía quedan algunos aspectos a resolver, especialmente aquéllos relacionados con la calidad de los servicios ofrecidos.
En la presente charla, nos centraremos en el soporte a servicios a tres niveles diferentes: i) a nivel de capa física y de enlace, ii) a nivel de capa de red, y iii) a nivel de capa de transporte. En concreto, se presentarán algunos modelos de propagación y de atenuación para los estándares 802.11p y LTE, un protocolo de enrutamiento híbrido llamado AVE y un protocolo de entrega de contenidos para canales con pérdidas.


Seminario: Simulación Avanzada de Redes Vehiculares

Ponente: Miguel Báguena

Día: miércoles 21 de diciembre

Hora: 10:00

Lugar: Laboratorio de Telemática (EUPT, Teruel)

Duración: 2 horas

Abstract

Debido al alto coste de realizar pruebas utilizando despliegues reales a la hora de trabajar con Redes Vehiculares, una de las técnicas más utilizadas es la simulación. Esta simulación, como es de suponer, debe contemplar todo tipo de detalles, siempre dirigidos a incrementar el nivel de realismo de dichas simulaciones, así como la precisión de los resultados obtenidos. En el presente seminario, se presentarán los detalles más importantes a tener en cuenta a la hora de realizar simulaciones, se dará un tutorial del simulador OMNeT++ y de sus modelos para simulación de entornos vehiculares, y finalmente se presentará la herramienta VACAMOBIL, que permite la simulación de movilidad realista de vehículos en entornos metropolitanos en base a parámetros predefinidos.


Seminario: "Image Restoration in Participating Media"

Título: Image Restoration in Participating Media and Its Application For Robotics

Ponente: Paulo Drews-Jr

Día: Martes 29 de noviembre

Hora: 12:00

Lugar: Seminario del Departamento de Informática e Ingeniería de Sistemas

Abstract: The computer vision and image processing fields have witnessed remarkable advances on a wide range of problems, most of them with direct impact on the robotics field. Regardless of the advances, they assumed that the light rays travel through the medium without any alteration. However, the medium can change the intensity and the direction of the light rays being named participating medium. Among them, the most important in terms of robotics application and discussed here are the turbid water and haze/fog. This talk presents the adopted approaches to reduce the effects generated by the medium. They are based on a physical light attenuation model which is used to enhance the quality of visual features of the images and the applicability of traditional computer vision techniques for these images. Furthermore, the talk shows the recent works developed in NAUTEC/FURG on robotics and computer vision.

Short Bio: Paulo Drews-Jr is a D.Sc. and M.Sc. in Computer Science at Federal University of Minas Gerais (UFMG) under supervisor of Prof. Mario Campos, Belo Horizonte, Brazil. His doctoral dissertation was focused in restoration of images acquired in participating media. His master thesis was focused on change detection and shape retrieval in 3D Maps. He graduated in Computer Engineering at the Federal Univesity of Rio Grande (FURG), Rio Grande, Brazil. His main research interests are robotics, computer vision, image processing, pattern recognition and machine learning. He was a researcher at the ISR Coimbra's Mobile Robotics Lab, collaborating, as a research fellow on the FP projects Prometheus and IRPS. He was also a visiting researcher in the Autonomous System Lab at QCAT-CSIRO, Brisbane, Australia. Currently, he is Assistant Professor at Federal University of Rio Grande.


Seminario: Analysis and control of concurrent systems

Ponentes:

- Gaiyun Liu. Xidian University. Xidian University. Xi’an, China “Robust Supervisory Control of Automated Manufacturing Systems”

- Kamel Barkaoui. Conservatoire National des Arts et Métiers (Le Cnam- Paris) “Maximal persistent steps for reduced state space generation”

- ZhiWu Li. Xidian University. Xi’an, China “Maximally Permissive Petri Net Supervisor with Novel Structures”

Día: 21 de noviembre de 2016

Hora: 12h

Lugar: Seminario (bloque 5, 2ª planta) del Instituto

Universitario de Investigación en Ingeniería de Aragón (I3A).

Edificio I+D+I – Campus Río Ebro. C/Mariano Esquillor s/n.

50018, Zaragoza

Duración estimada: 2 horas.

Abstracts and Bios:

- Gaiyun Liu received the BS and the first PhD degrees from Xidian University, Xi’an, China, in 2006 and 2011, respectively. She received the second PhD degree from Conservatoire National des Arts et Métiers (Le Cnam), Paris, France, in 2015. Since 2011, she has been with Xidian University, where she is currently an associate professor with the School of Electro-Mechanical Engineering. Her research interests include Petri net theory and applications and supervisory control of discrete event systems. She published 17 papers in international journals and conferences.

Title: Robust Supervisory Control of Automated Manufacturing Systems

Abstract: A variety of deadlock control policies based on Petri nets have been proposed for automated manufacturing systems (AMSs). Almost all of them assume that resources do not fail. Actually, resource failures are usually inevitable in real-world systems, which pose challenges in supervisory control of contemporary AMSs. When resources are unreliable, it is infeasible or impossible to apply the existing control strategies and failure-derived deadlocks in the disturbed system may be caused. Reanalysis of the disturbed system is usually necessary, which may lead to high cost of production. For a class of Petri nets (systems of simple sequential processes with resources, S3PR), this talk introduces a robust deadlock control policy in which the factor of resource failures is considered. Recovery subnets and monitors are designed for unreliable resources and strict minimal siphons that may be emptied, respectively. Normal and inhibitor arcs are used to connect monitors with recovery subnets in case of necessity. Then reanalysis of the original Petri net is avoided and a robust liveness-enforcing supervisor is derived.

- Kamel Barkaoui is full professor at the Department of Computer Science of Conservatoire National des Arts et Métiers (Le Cnam - Paris) since 2002. He holds a Ph.D in Computer Science (1988) and Habilitation à Diriger des Recherches (1998) from Université Paris 6 (UPMC). His research interests include formal methods for specification, verification, control and performance evaluation of concurrent and distributed systems. He published over 100 peer-reviewed papers in international journals and conferences, supervised 28 PhD theses, co-edited 10 books. He received the 1995 IEEE Int. Conf. on System Man and Cybernetics Outstanding Paper Award and Recipient of the Prime d’Excellence Scientifique and of the PEDR since 1998. He was PC co-chair of the 3rd International Colloquium on Theoretical Aspects of Computing (ICTAC2006), General co-chair of the 18th International Symposium on Formal Methods (FM2012) and General chair of the 35th International Conference on Application and Theory of Petri Nets and Concurrency (Petri Nets2014) and of the 14th International Conference on Application of Concurrency to System Design (ACSD'2014). He was a Guest Editor of Formal Aspects of Computing Journal (FACJ), of Journal of Systems and Software (JSS) and ACM Transactions on Embedded Computing Systems (TECS) and serves as Associate Editor for the International Journal of Critical Computer-Based Systems (IJCCBS). He is the founder and SC chair of the International Conference on Verification and Evaluation of Computer and Communication Systems (VECoS).

Title: Maximal persistent steps for reduced state space generation

Abstract: The state explosion problem is the main obstacle for the formal verification of concurrent systems, as they are generally based on an interleaving semantics, where all possible firing orders of concurrent actions are exhaustively explored. The step graph method aims at attenuating the state explosion problem, by executing, in one step, the largest possible number of concurrent transitions. The idea is to represent, by a single path, the largest possible set of equivalent firing sequences of the model. All transitions of the sequences are represented in the path but the concurrent ones are grouped together in steps. However, even if all transitions within the same step are concurrent, their firing in one step may prevent enabledness of some conflicting transitions (deferred conflicts) that are enabled, when the transitions of the step are fired in specific orders. Consequently, some equivalent maximal firing sequences will not be represented in the step graph. The step graph method, proposed in the literature for Petri nets, handles such a problem by systematically excluding from every step transitions that are in weak conflict relation. In this talk we present some recent results that significantly alleviate this limitation. After introducing the notion of persistent firing steps, we propose a maximal persistent step graph (MPSG) method and show that all and only all non-equivalent maximal firing sequences of the model are represented in the MPSG. Afterwards, we establish a practical sufficient condition for a step to be persistent that yields a significant reduction of the state space, especially for large asynchronous concurrent systems.

- ZhiWu Li received the B.S., M.S., and Ph.D. degrees from Xidian University in 1989, 1992, and 1995, respectively. He joined Xidian University in 1992. His interests include discrete event systems and Pertri nets. He published two monographs in Springer and CRC Press and 70+ papers in Automatica and IEEE Transactions (mostly regular). His work was cited by engineers and researchers from more than 40 countries and areas, including prestigious R&D institutes such as IBM, Volvo, HP, GE, GM, ABB. He serves (served) an Associate Editor of the IEEE Trans. Automation Science and Engineering, IEEE Trans. Systems, Man, and Cybernetics, Part A: Systems and Human Beings, IEEE Trans. Systems, Man, and Cybernetics: Systems, and Information Sciences (Elsevier). Dr. Li is a recipient of Alexander von Humboldt Research Grant and Research in Paris. He was selected in 2014, 2015, and 2016 Thomson Reuters Highly Cited Researchers in the category of Engineering. He is a Fellow of IEEE.

Title: Maximally Permissive Petri Net Supervisor with Novel Structures

Abstract: This talk introduces two classes of inhibitor arcs: interval inhibitor arcs and data inhibitor arcs that are used to design a maximally permissive liveness-enforcing supervisor for an automated manufacturing system modeled with Petri nets. The development of the novel inhibitor arcs is motivated by the fact that there usually does not exist an optimal supervisor represented by pure Petri nets when the legal space is non-convex. A significant result derived from this new net structure is that the pertinent approach can always lead to an optimal supervisor with only one control place for bounded Petri nets on the premise that such a supervisor exists.

Seminario: Federation Management (Cloud and Otherwise) Using the OpenStack Mitaka Release

Día: 28 de octubre de 2016

Hora: 12:00

Lugar: Seminario del Instituto Universitario de Investigación en Ingeniería de Aragón (I3A)

Edificio I+D+I – Campus Río Ebro

C/Mariano Esquillor s/n

50018, Zaragoza

Invited Talk: Dr. Craig Alexander Lee

Title: “Federation Management (Cloud and Otherwise) Using the OpenStack Mitaka Release”

Abstract:

Federation is a fundamental goal of many distributed systems, including clouds, and more accurately, services. As a possible corollary to Metcalfe's Law, the value of a cloud is proportional to the square of the number of users and services connected. A key concept to grasp is that in a service architecture, federation can be done at any level in the system stack. A cloud is simply a set of services, and a SaaS cloud can include any arbitrary, application-level services. As such, cloud federation is a special case of general service federation.

This talk establishes the basic requirements and properties for federation and reviews relevant, existing standards. We review the NIST Cloud Deployment Models to extract the more fundamental properties of _trust boundaries_ and _trust topologies_. This allows us to describe a set of Federation Deployment Models.

Armed with this clear understanding of the federation design space, we describe our current work in building a prototype, KeyVOMS, a Keystone-based Virtual Organization Management System. KeyVOMS uses the Virtual Organization concept (originating in the grid computing era) for managing federations, and is based on a re-purposed OpenStack Keystone v3 server. Our current implementation is a simple, centralized, third-party server, based on the Havana release. We are currently implementing a Peer-to-Peer version based on the Mitaka release that will leverage the latest Keystone capabilities. We conclude by reviewing related work and future work. Biographical Sketch: Craig Alexander Lee Craig A. Lee is a Senior Scientist in the Computer Systems Research Department of The Aerospace Corporation, a non-profit, federally funded, research and development center. Dr. Lee has worked in the area of parallel and distributed computing for the last 35 years. This work led to Dr. Lee's involvement in the Open Grid Forum (OGF) where he is served as President from 2007 to 2010. In this capacity, he served as the point of contact with DMTF, SNIA, TMF, the Open Cloud Consortium, Cloud Security Alliance, OMG, OASIS, and other organizations working on cloud standards. During this period, OGF produced the Open Cloud Computing Interface (OCCI), an open standard API for infrastructure clouds. Dr. Lee is now supporting the National Institute for Standards and Technology (NIST) in their Cloud Computing Program. He contributed significantly to the NIST Cloud Technology Roadmap and is now supporting the NIST Cloud Federation Working Group. Dr. Lee is on the steering committee for the ACM/IEEE International Conference on Utility and Cloud Computing (UCC) and the ACM/IEEE International Symposium on Cluster, Cloud and Grid Computing (CCGrid). He has served on the program committee for many other conferences and workshops, and has served as a panelist for the NSF, NASA, DOE, and as an international evaluator for INRIA. Dr. Lee has published over 80 technical works, including four book chapters and seven edited volumes and issues, and is an associate editor of Future Generation Computing Systems (Elsevier) and on the editorial board of the International Journal of Cloud Computing (Inderscience). He is also a member of the ACM and the IEEE Computer Society. Dr. Lee has a Ph.D. in Computer Science from the University of California, Irvine, and has lectured undergraduate computer science courses at UCLA.


Seminario: Modeling and Analysis of Industrial Robots using Petri nets

Lugar: Seminario Departamento

Día: 20 septiembre

Hora: 10:30 horas

Ponente: Carlos Pascal

Technical University of Iasi, Romania

Abstract:

The presentation contains a brief overview of systems and computer engineering research and laboratories at Technical University of Iasi: sound of vision project [1], autonomous mobile robots - sample gathering project [2], and other small projects. This is followed by my research on holonic manufacturing systems [3] and [4]. Here, it is discussed about connecting the implementation of manufacturing systems with their behavioral models that can be obtained by the means of colored Petri nets [5]. Distributed algorithms are between the topical approaches used in manufacturing control architectures. Their validation and evaluation techniques are challengeable due the non-deterministic execution and the partial knowledge of each entity. The development phase can hide several issues; for example, to reproduce a failure or worse execution based on the simulation data may be difficult. With respect to this, one possibility is to develop a model-prototype enclosing some basic characteristics: simple and understandable, shareable, modifiable, catching the partial/entire system’s behavior, allowing space state analysis and simulations. Such a model can be built up through colored Petri nets, using the CPN Tool [5]. The considered example regards the model for three distributed constraint satisfaction algorithms [6]: synchronous backtracking, asynchronous backtracking and weak-commitment search. Following this approach, other distributed protocols can be designed and validated, too.

1. www.soundofvision.net

2. www.loris.tuiasi.ro

3. Panescu, D., Pascal, C., 2012. HAPBA – a Holonic Adaptive Plan-Based Architecture. In: Borangiu, Th., Thomas, A., Trentesaux, D. (eds.) Service Orientation in Holonic and Multi-Agent Manufacturing Control, Springer, pp. 61-74.

4. Panescu, D., Pascal, C., 2015. Holonic coordination obtained by joining the contract net protocol with constraint satisfaction. Computers in Industry, vol. 81, pp. 32-46.

5. Jensen, K., Kristensen, L.M., 2009. Coloured Petri nets: modeling and validation of concurrent systems. SpringerVerlag, New York.

6. Yokoo, M., Durfee, E., Ishida, T., Kuwabara, K., 1998. The distributed constraint satisfaction problem: formalization and algorithms. IEEE Transaction on Knowledge and Data Engineering, 10(5), pp. 673-685.


Seminario: Stochastic models for operating room planning and scheduling

Lugar: Seminario Departamento

Día: 13 septiembre

Hora: 16 horas

Ponente: Xiaolan XIE

Centre for Biomedical & Healthcare Engineering
CNRS UMR 6158 LIMOS
Ecole Nationale Supérieure des Mines de Saint Etienne
158 cours Fauriel, 42023 Saint Etienne
xie@emse.fr

and

Centre for Healthcare Engineering
Dept. Industrial Engineering & Management
Shanghai Jiao Tong University
800 Dong Chuan Rd., Shanghai 200240, P.R. China

Abstract:

In this talk, I will first give an overview of my research in healthcare in France and in China since over 10 years. I will then focus on stochastic models for planning and dynamic scheduling of operating rooms.

I first present stochastic models for Operating Room (OR) planning with two classes of patients: elective surgery and emergency surgery. Elective cases can be planned starting from an earliest date with a patient related cost depending on the surgery date. Emergency cases arrive randomly and have to be performed in the day of arrival. The planning problem consists in assign elective cases to different periods over a planning horizon in order to minimize the sum of elective patient related costs and overtime costs of operating rooms. We first propose a basic stochastic mathematical programming model and a Monte Carlo optimization method combining the Monte Carlo simulation and Mixed Integer Programming. The convergence to real optimal and properties of the convergence rate as the computation budget increase are established. We then extended the basic OR planning model to include important factors such as patient to OR assignment, random operating times and overtime capacity constraints. The Monte Carlo optimization method can be extended to solve the more general case but the computation time quickly goes beyond acceptable limit. We then propose a column generation approach that is able to provide efficient solutions for problems of realistic size with duality gap of less than 1%.

I will then address the daily surgery scheduling in a multi-OR setting with random surgery durations. We first address the dynamic assignment of a given set of surgeries with planned surgeon arrival times also called appointment times. Surgeries are assigned dynamically to ORs. The goal is to minimize the total expected cost incurred by surgeon waiting and OR idling/overtime. We first formulate the problem as a multi-stage stochastic programming model. An efficient algorithm is then proposed by combining a two-stage stochastic programming approximation and some look-ahead strategies. A perfect information-based lower bound of the optimal expected cost is given to evaluate the optimality gap of the dynamic assignment strategy. Numerical results show that the dynamic scheduling and optimization with the proposed approach significantly improve the performance of usual static scheduling and First Come First Serve (FCFS) strategies.

We then address the optimization of surgeon appointment times for a sequence of surgeries. Surgeries are assigned to ORs dynamically on a FCFS basis. It materially differs from past literature in the sense that dynamic assignments are proactively anticipated in the determination of appointment times. A discrete-event framework is proposed to model the execution of the surgery schedule and to evaluate the sample path gradient of a total cost incurred by surgeon waiting, OR idling and OR overtime. The sample path cost function is shown to be unimodal, Lipchitz continuous and differentiable w.p.1 and the expected cost function continuously differentiable. A stochastic approximation algorithm based on unbiased gradient estimators is proposed and extensive numerical experiments suggest that it converges to a global optimum. A series of numerical experiments are performed to show the significant benefits of Multi-OR and properties of the optimal solution with respect to various system parameters such as cost structure and numbers of surgeries and ORs.

The last part of this talk is devoted to the daily overtime management to provide OR-teams with different degrees of guarantee of on-time end of the day. This problem is modelled as two stochastic bin packing problems with chance constraints: a stochastic programming model and a robust formulation with moment information. We present a tailored branch-and-price approach based on a column generation reformulation for which CVaR and probabilistic sets are employed to speed-up the resolution of the pricing problems. For the distributionally robust formulation the worst-case distribution is shown to be a three-point distribution, analytical expression of the worst chance probability derived and simple mixed integer linear program model obtained. We implement a series of real-data-based numerical experiments to show that the benefits of optimal allocation and illustrate the value of robust solution.


Seminario: Evaluación de la calidad de objetos de aprendizaje y cursos online a través de estándares y metadatos

Lugar: Seminario DIIS Ada Byron
Fecha: miércoles 16 de marzo de 2016
Hora: 16 horas

Ponente
Daniel Pons Betrián es Ingeniero en Informática por la Universidad de Zaragoza y Doctor por la Universidad de Alcalá. Ha trabajado como ingeniero informático en Telefónica de España S.A.U.. Es profesor de informática en el cuerpo de Profesores Técnicos de Formación Profesional en la especialidad de Sistemas y Aplicaciones Informáticas y actualmente trabaja como Asesor Técnico del Centro Aragonés de Tecnologías para la Educación. Está colaborando con el subcomité de AENOR “AEN/CTN71/SC36 Tecnologías de la Información para el Aprendizaje” para el desarrollo de una norma española sobre la calidad de los recursos educativos.

Resumen
El crecimiento exponencial de los recursos educativos digitales lleva a estudiar la aplicación de indicadores para la medición de la calidad de objetos de aprendizaje y de la formación virtual. Los metadatos son un lugar idóneo para almacenar información sobre la calidad de los objetos de aprendizaje. La estructura de metadatos para el e-learning denominada LQM, basada en estándares de metadatos de objetos de aprendizaje y en estándares de calidad de la formación virtual, permite un mejor desempeño para los sistemas de catalogación, almacenamiento y localización de recursos educativos utilizando la calidad como criterio de búsqueda. Se presentarán los principios de diseño de una herramienta software de catalogación de objetos de aprendizaje, la cual realiza una evaluación automática de la calidad de la formación virtual.


Charla: Avances recientes en redes vehiculares

Ponente: Manuel Fogué Cortés
Lugar: Laboratoria de Telemática de la EUPT
Día: 5 de febrero de 2016
Hora: 9:00 AM
Duración aproximada: 1.5 horas

Resumen
La integración de las tecnologías de la telecomunicación a la industria automovilística es un hecho que cambiará, en un plazo de tiempo relativamente corto, la forma en que entendemos el transporte. Un enorme abanico de aplicaciones relativas a la seguridad y al confort de los pasajeros son posibles en este entorno, pero surgen numerosos retos a resolver, desde el desarrollo de simuladores que reduzcan el coste de experimentación, a la obtención de algoritmos inteligentes que sean capaces de procesar la información recogida por el gran número de vehículos. En la charla se presentarán soluciones desarrolladas durante los últimos años, así como retos a solucionar en el futuro.

Breve bio del ponente
Manuel Fogué es profesor ayudante doctor en la Universidad de Zaragoza. Estudió Ingeniería Técnica en Informática de Gestión en la Universidad de Zaragoza (2007), Ingeniería Informática en la Universitat Jaume I de Castellón (2009), y obtuvo su Doctorado de nuevo en la Universidad de Zaragoza en 2012. Es miembro del grupo de investigación Intelligent Networks and Information Technologies (iNiT). Sus intereses de investigación incluyen la simulación de VANETs, los Sistemas de Transporte Inteligente, las comunicaciones vehículo a vehículo (V2V) y vehículo a infraestructura (V2I), así como la aplicación de Inteligencia Artificial a los entornos vehiculares.


Charla: Vehicular Challenges

Ponente: Julio A. Sangüesa
Lugar: Laboratoria de Telemática de la EUPT
Día: 5 de febrero de 2016
Hora: 11:00 AM
Duración aproximada: 1.5 horas

Resumen
Debido al desarrollo tecnológico, el coche inteligente se está convirtiendo en uno de los protagonistas indiscutibles del siglo XXI. En los últimos años han sido numerosos los avances que vienen siendo introducidos paulatinamente en nuestros vehiculos. Sin embargo, todavía existen dos factores clave en el vehículo del futuro, por un lado el combustible que dichos vehículos consumirán, siendo el vehículo eléctrico uno de los más prometedores, y por otro lado los sistemas de comunicación entre vehículos, que en un futuro a medio plazo podrán permitir la existencia del vehículo autónomo y la conducción cooperativa. Por otra parte, uno de los principales hándicaps de este sector es la lenta renovación del parque móvil existente, por lo que se hace necesaria la incorporación de tecnologías intermedias que nos permitan acelerar la introducción progresiva de las nuevas propuestas y servicios.

Breve bio del ponente
Julio A. Sangüesa es investigador en el grupo de Redes de Computadores de la Universitat Politècnica de València y colaborador del grupo iNiT. Es Ingeniero Técnico en Informática de Gestión por la Universidad de Zaragoza; posteriormente realizó el Grado en Ingeniería en Sistemas de Información por la Universidad Católica de Ávila y un Máster en Software Libre en la Universidad Oberta de Catalunya. En 2014 obtuvo el grado de Doctor en la Universidad de Zaragoza. Sus intereses de investigación incluyen la simulación de VANETs, Vehículos Eléctricos, Sistemas de Transporte Inteligente, y las comunicaciones vehículo a vehículo (V2V).

Seminario: Bayesian Networks for dependability and vulnerability assessment

Ponente: Stefano Marrone de la Universitá di Napoli Federico Secondo
Lugar: Seminario del DIIS
Fecha: 3/2/2016
Hora: 15:30

Resumen:
Reliability assessment has been traditionally dominated by the Fault Trees vs Markov Chains discussion. Between this two extremes, Bayesian Networks are a valid alternative option since a greater modeling power than the Fault Trees and greater solution efficiency than Markov Chains. This talk introduces such language showing how it can be used in the modeling of reliability, safety and security of complex and critical systems. The talk also summarizes some experiences of framing the BN formalism into model-driven approaches.

Charla: Avances en percepción robótica con cámaras 3D

Ponente: Javier González-Jiménez
Hora: 9:00AM
Duración aproximada: 1 hora
Lugar: Seminario del Departamento de Informática e Ingeniería de Sistemas.

Abstract:
La aparición reciente de cámaras de profundidad compactas y de bajo coste ha supuesto un revulsivo en la capacidad sensorial de los robots autónomos, hasta el punto que estamos asistiendo a avances que hacen presagiar a muy corto plazo aplicaciones robóticas que hasta hace pocos años se pensaban lejanas en el tiempo. El aprovechamiento en robótica de estas nuevas tecnologías (cámaras TOF y RGB¬D, fundamentalmente) exige métodos y algoritmos específicos que procesen eficientemente la ingente cantidad de datos suministrados y, en última instancia, proporcionen al robot información del entorno útil para navegar, tomar decisiones, interaccionar con personas, etc. En esta charla, el prof. González¬Jiménez revisa las aportaciones del grupo MAPIR¬UMA (http://mapir.isa.uma.es) en este campo. Entre otros, se presentan resultados en odometría visual y su generalización al flujo de velocidad de los puntos de la escena, la construcción de mapas basados en planos, el reconocimiento de objetos y habitaciones basado en información contextual, y la calibración extrínseca relativa de varios sensores a bordo del robot.


Charla: Visual Indoor Localization and Navigation

Ponente: Eckehard Steinbach
Hora: 10:00AM
Duración aproximada: 1 hora
Lugar: Seminario del Departamento de Informática e Ingeniería de Sistemas.

Abstract:
This talk addresses the challenge of localizing mobile devices inside large building structures with a special focus on visual localization. In visual indoor localization, the position and orientation of a mobile device are recovered using the images captured by its built-in video camera. Given these images, location information can be obtained by comparing the captured images with a data base of geo-tagged reference images of the environment using concepts from the field of content-based image retrieval. Flexible and cost-efficient acquisition of this reference image data base is a key technology in this context. Hence, a significant part of the talk will be devoted to the acquisition of this data base for a previously unknown environment. Once this reference data is available, then pose estimation needs to be performed with minimum latency, and hence the talk will also cover concepts for minimizing the client server communication for visual localization. In order to improve the robustness and accuracy of the indoor localization process, the visual information needs to be fused with other sources of information, such as pedometer, WiFi, magnetometer, etc. To this end we present a graph-based fusion algorithm which substantially reduces the complexity of a particle filter-based sensor fusion approach and makes it applicable on standard mobile devices. The talk will end with a demonstration of our prototype visual indoor localization and navigation system and an overview of the remaining challenges on the way to GPS-free indoor navigation.



Seminario Investigación: 3D Object Detection for Visual Scene Understanding

Título:"3D Object Detection for Visual Scene Understanding”.
Ponente: Roberto Lopez Sastre.
Día: 13 de Enero.
Lugar: Seminario del DIIS.
Hora: 16.00h.

Abstract:
Inspired by the ability of humans to interpret 3D scenes nearly effortlessly, the problem of 3D scene understanding has long been advocated as the "holy grail" of computer vision. Recently, there has been considerable progress on many sub-taks of the overall 3D scene understanding problem.
In this talk we will highlight progress on one of these essential sub-problems of scene understanding such as simultaneous object detection and pose estimation. We will report on our current attempt towards 3D object detection models using Hough Forest (HF). In particular, we will introduce two approaches:
1) the HF with Probabilistic Enhanced Voting; and
2) the Boosted Hough Forest model.

Más información «


Seminario transversal: Publish or Perish, Part 1: Why, When, Where, How much?

Ponente: Juan Domingo Tardós, Dept. Informática e Ingeniería de Sistemas,Universidad de Zaragoza
Dia: Lunes, 14 de Diciembre de 2015
Hora:16:00-18:00h
Lugar: Aula A12 Edificio Ada Byron


Seminario transversal: Publish or Perish, Part 2: How?

Ponente: Juan Domingo Tardós, Dept. Informática e Ingeniería de Sistemas,Universidad de Zaragoza
Dia: Miércoles, 16 de Diciembre de 2015
Hora:16:00-18:00h
Lugar: Aula A12 Edificio Ada Byron


Seminario transversal:Manage your research career. Life after PhD

Ponente: Luis Montesano, Dept. Informática e Ingeniería de Sistemas, Universidad de Zaragoza
Dia: Jueves, 17 de Diciembre de 2015
Hora:16:00-17:30h
Lugar: Aula A12 Edificio Ada Byron

Seminario transversal:How to give a good research talk

Ponente: Diego Gutiérrez, Dept. Informática e Ingeniería de Sistemas,Universidad de Zaragoza
Dia: Jueves, 10 de Diciembre de 2015
Hora:10:00-11:30h
Lugar: Aula A12 Edificio Ada Byron


Seminario Investigación: Planificación y ejecución de proyectos de manufactura

Ponente: Antonio de la Vega, director durante los últimos años del del Departamento de Ingeniería de General Motors España en Figueruelas
Jueves 26 de noviembre, de 10:00-12:00, en el Aula A-14 del Ada Byron
Título: Planificación y ejecución de proyectos de manufactura

El objetivo del seminario es transmitir la experiencia durante los últimos años en el desarrollo de proyectos de manufactura. Se explicará una metodología para la planificación, implantación y seguimiento de Proyectos de manufactura. Se plantearán todas las fases de un proyecto, y se ilustrará con ejemplos prácticos, herramientas y medios de planificación y control. Se expondrán cambios tecnológicos llevados a cabo en sucesivos proyectos a lo largo del tiempo, así como motivos para realizarlos y las ventajas esperadas y obtenidas. Estos cambios están relacionados con la implantación de automatismos, de robots y de sensores, en especial visión artificial, en diferentes etapas del proceso de fabricación.Se discutirán las experiencias adquiridas durante 18 años de trabajo en este campo.


Charla: Time-lapse Mining from Internet Photos

Jueves, 24 Sep. 12:00. Seminario Dpto.
Nombre: Ricardo Martin Brualla, PhD Candidate at University of Washington Título: Time-lapse Mining from Internet Photos.

Abstract:
I will describe an approach for synthesizing time-lapse videos of popular landmarks from large community photo collections. The approach is completely automated and leverages the vast quantity of photos available online. First, we cluster 86 million photos into landmarks and popular viewpoints. Then, we sort the photos by date and warp each photo onto a common viewpoint. Finally, we stabilize the appearance of the sequence to compensate for lighting effects and minimize flicker. The approach is also extended to moving camera paths where a virtual camera moves both in space and time, creating time-lapse videos with compelling parallax effects. Our resulting time-lapses show diverse changes in the world's most popular sites, like glaciers shrinking, skyscrapers being constructed, and waterfalls changing course

(see this video ).



Seminario Investigación: Constraint Algorithms for Molecular Dynamics

Ponente: Pablo García Risueño Institut für Physik und IRIS Adlershof, Humboldt Universität zu Berlin
Martes 7 de julio, 12:00 en el Seminario 23 del Ada Byron
Título: Constraint Algorithms for Molecular Dynamics

Molecular Dynamics simulations of biological molecules have a central role in science at present. In order to simulate relevant phenomena, it is important to increase the time step of the simulations. At this talk we will summarize a research project that is currently carried out at the University of Zaragoza, and whose goal is to increase the time step of these simulations.


Seminario de Investigación:Estudios sobre el comportamiento de tráfico Ethernet y Wi-Fi

Ponente: Mg. Ing. Santiago Pérez, Universidad Tecnológica Nacional de Argentina.
Dia y hora: martes 07 de julio de 10:00 a 12:00 h. (incluyendo tiempo para preguntas/discusión).
Lugar: Seminario del Departamento de Informática e Ingeniería de Sistemas. Edificio Ada Byron, planta 1ª. Campus Río Ebro.

Resumen:
El seminario se divide en dos partes, y se refiere a líneas abiertas y trabajos recientes de investigación sobre el comportamiento de tráfico de red en ciertas condiciones particulares.
1. Modelación del tráfico de video Ethernet. Estudios experimentales de su comportamiento en redes cableadas Ethernet.
Se plantean los estudios experimentales realizados sobre la conducta del tráfico de video comprimido en redes cableadas, y que a mediano plazo, incluirán también el mismo estudio sobre las redes Wi-Fi. Se presentará la función de los codecs de video, y la metodología utilizada para el estudio de las capturas del tráfico generado con un video de estudio.
Se incluirán los siguientes tópicos: ingeniería de tráfico, avances alcanzados, video digital, compresión de video, formato de la trama de video, escenarios de prueba, algunos resultados, modelos de tráfico de video.
Al finalizar esta primera parte, se resumirá un trabajo presentado en un Congreso IEEE.

2. Impacto de BER (Bit Error Rate) en el rendimiento de redes Inalámbricas Wi-Fi. Estudios por simulación. Se plantean los estudios de simulación realizados sobre la conducta del tráfico de Wi-Fi en presencia de interferencias. A partir del análisis de escenarios simulados (usando Redes de Petri), se expondrán estudios de máximo rendimiento Wi-Fi posibles para distintas longitudes de tramas. Y se presentará el modelo de error utilizado, y los avances de estudios variando el nivel de error (BER), y su impacto sobre los rendimientos máximos.
Se incluirán los siguientes tópicos: máximo rendimiento en Wi-Fi, modelo de error de Gilbert-Elliot, escenarios de prueba, demostración, conclusiones.


Conferencia: Proyecto Romol: Runtime Aware Architectures

Titulo Conferencia: Proyecto Romol: Runtime Aware Architectures
Fecha: Lunes 1 de Junio, 18h
Lugar: Aula A03, Edificio Ada Byron, Campus Rio Ebro

Abstract:
In the last years the traditional ways to keep the increase of hardware performance to the rate predicted by the Moore's Law vanished. When uni-cores were the norm, hardware design was decoupled from the software stack thanks to a well defined Instruction Set Architecture (ISA). This simple interface allowed developing applications without worrying too much about the underlying hardware, while computer architects proposed techniques to aggressively exploit Instruction-Level Parallelism (ILP) in superscalar processors. Current multi-cores are designed as simple symmetric multiprocessors on a chip. While these designs are able to compensate the clock frequency stagnation, they face multiple problems in terms of power consumption, programmability, resilience or memory. The solution is to give more responsibility to the runtime system and to let it tightly collaborate with the hardware. The runtime has to drive the design of future multi-cores architectures. In this talk, we introduce an approach towards a Runtime-Aware Architecture (RAA), a massively parallel architecture designed from the runtime's perspective.

CV: Mateo Valero Cortés.
http://www.bsc.es/cv-mateo
Doctor Ingeniero de Telecomunicación. Desde 1983, es catedrático de la Universidad Politécnica de Cataluña (UPC). Ha publicado más de 600 artículos en el área de la arquitectura de los computadores de altas prestaciones. Director del Barcelona Supercomputing Center – Centro Nacional de Supercomputación. Entre sus premios, el Premio Eckert-Mauchly 2007 que es el más importante a nivel mundial en el ámbito de la Arquitectura de Computadores, Premio “Harry H. Goode” 2009 otorgado por el IEEE, Premio ACM distinguished Service Award 2013 otorgado por el ACM, dos premios nacionales de investigación que son: el "Julio Rey Pastor" en Informática y Matemáticas y el “Leonardo Torres Quevedo” en Ingeniería. Premio Rey Jaime I de Investigación de la Generalitat Valenciana. “Hall of Fame” en el marco del ICT European Program, seleccionado como uno de los 25 investigadores europeos más influyentes en IT, Tecnologías de la Información. Fellow del IEEE, Fellow distinguido de Intel y Fellow del ACM. Es miembro fundacional de la Real Academia de Ingeniería de España, académico correspondiente de la Real Academia de Ciencias Exactas, Físicas y Naturales, miembro de la Real Academia de Ciencias y Artes de Barcelona, Miembro de la Academia Europea y académico correspondiente de la Academia Mexicana de Ciencias. Es Doctor Honoris Causa de las Universidades de Chalmers, Belgrado, Las Palmas, Veracruz, Zaragoza, Complutense de Madrid y Universidad de Cantabria en Santander. En 1998, fue elegido hijo predilecto de su pueblo, y en el año 2006, la asociación de madres y padres de alumnos de Alfamén, decidió poner su nombre al Colegio público donde el profesor Valero había estudiado.

Seminario de Investigación:Controlling Multiple Agents in their Pursuit of Multiple Objectives

Título: Controlling Multiple Agents in their Pursuit of Multiple Objectives
Ponente: Dusan M. Stipanovic
Associate Professor at University of Illinois at Urbana-Champaign http://ise.illinois.edu/directory/profile/dusan
Fecha y hora: Martes 19 mayo, 10:00
Lugar: Seminario 21 Edificio Ada Byron

Abstract: The challenges of controlling dynamic systems with multiple objectives are related to and furthermore include problems in multi-player dynamic games, multiobjective optimization, and decentralized control and estimation. The additional complexity is introduced through nonlinear dynamic models with possible delays and perturbations as well as various state, input and communication constraints. In this talk we will present a number of results related to control and coordination of multi-agent systems pursuing multiple objectives. To illustrate these results some particular examples of controlling multiple agents pursuing multiple objectives such as guaranteed capture or evasion, collision avoidance, coverage control, proximity, and tracking, will be presented.

Seminario de Investigación: Planificación y Ejecución de Proyectos de Manufactura

Ponente: D. Antonio de la Vega

En los últimos años ha sido Director del Departamento de Ingeniería de General Motors España en Figueruelas. Es Ingeniero Industrial por la Escuela Técnica Superior de Ingenieros Industriales de Madrid. Desde 1980 ha trabajado en General Motors España (Figueruelas), con responsabilidad en áreas de mantenimiento, energía, medio ambiente, Ingeniería, control de calidad, producción simplificada y trabajo en equipo.

Seminario 2: 15 de mayo de 12 a 14h. en el Aula A05 del Ada Byron

Resumen
El objetivo del seminario es transmitir la experiencia durante los últimos años en el desarrollo de proyectos de manufactura. Se explicará una metodología para la planificación, implantación y seguimiento de Proyectos de manufactura. Se plantearán todas las fases de un proyecto, y se ilustrará con ejemplos prácticos, herramientas y medios de planificación y control. Se expondrán cambios tecnológicos llevados a cabo en sucesivos proyectos a lo largo del tiempo, así como motivos para realizarlos y las ventajas esperadas y obtenidas. Estos cambios están relacionados con la implantación de automatismos, de robots y de sensores, en especial visión artificial, en diferentes etapas del proceso de fabricación.Se discutirán las experiencias adquiridas durante 18 años de trabajo en este campo.

Seminario de Investigación: Planificación y Ejecución de Proyectos de Manufactura

Ponente: D. Antonio de la Vega

En los últimos años ha sido Director del Departamento de Ingeniería de General Motors España en Figueruelas. Es Ingeniero Industrial por la Escuela Técnica Superior de Ingenieros Industriales de Madrid. Desde 1980 ha trabajado en General Motors España (Figueruelas), con responsabilidad en áreas de mantenimiento, energía, medio ambiente, Ingeniería, control de calidad, producción simplificada y trabajo en equipo.

Seminario 1: 8 de mayo de 12 a 14h. en el Salón de Actos del Ada Byron

Resumen
El objetivo del seminario es transmitir la experiencia durante los últimos años en el desarrollo de proyectos de manufactura. Se explicará una metodología para la planificación, implantación y seguimiento de Proyectos de manufactura. Se plantearán todas las fases de un proyecto, y se ilustrará con ejemplos prácticos, herramientas y medios de planificación y control. Se expondrán cambios tecnológicos llevados a cabo en sucesivos proyectos a lo largo del tiempo, así como motivos para realizarlos y las ventajas esperadas y obtenidas. Estos cambios están relacionados con la implantación de automatismos, de robots y de sensores, en especial visión artificial, en diferentes etapas del proceso de fabricación.Se discutirán las experiencias adquiridas durante 18 años de trabajo en este campo.

Seminario Investigación: High resolution seafloor surveying using Autonomous Underwater Vehicles

Ponente: Stefan Williams
Professor of Marine Robotics
Director of The Marine Systems Group
Head of Australia's Integrated Marine Observing System
Australian Centre for Field Robotics
The University of Sydney
http://www-personal.acfr.usyd.edu.au/stefanw/pmwiki/

Lugar: Salón de Actos, Edificio Ada Byron
Fecha y hora: Lunes 4 de Mayo de 2015, 12h

Abstract:
This talk will present insights gained from the operation of an Australia-wide benthic observing program designed to deliver precisely navigated, time series imagery of the seafloor using Autonomous Underwater Vehicles (AUVs). This initiative makes extensive use of AUVs to collect high-resolution stereo imagery, multibeam sonar and water column measurements on an annual or semi-annual basis at sites around Australia, spanning the full latitudinal range of the continent from tropical reefs in the north to temperate regions in the south. In this talk, we will provide an update on the status of this observing program and outline examples of research that has been facilitated by this work in the areas of environmental monitoring, archaeology and geosciences. Examples of these outcomes highlight our ability to revisit survey sites across multiple years and illustrate how automated tools can be used to identify patterns within the large volumes of data being collected.

Short Bio:
Stefan B. Williams is the Professor of Marine Robotics at the University of Sydney’s School of Aerospace, Mechanical and Mechatronic Engineering. He is a member of the Australian Centre for Field Robotics where he leads the Marine Robotics group. He is also the head of Australia’s Integrated Marine Observing System Autonomous Underwater Vehicle Facility. His research interests include Simultaneous Localisation and Mapping in unstructured underwater environments using visual and acoustic sensing, autonomous navigation and control and classification and clustering of large volumes of data collected by robotic systems. He received his PhD from the University of Sydney in 2002 and completed a Bachelor of Applied Science in Systems Design Engineering at the University of Waterloo, Canada in 1997.

Seminario Investigación: Image Features for Robust Registration

Ponente: Dr. Vincent Lepetit
Professor at the Institute for Computer Graphics and Vision. TU Graz.
http://cvlabwww.epfl.ch/~lepetit/
http://www.icg.tugraz.at/Members/lepetit/vincent-lepetits-homepage

Fecha y hora: Jueves 07 mayo 2015 a las 16:00
Lugar: Seminario S.22 Edificio Ada Byron

Abstract
Extracting image features such as feature points or edges is a critical step of many Computer Vision systems, however this is still performed with carefully handcrafted methods. In this talk, I will present a new Machine Learning-based approach to detecting local image features, with application to contour detection in natural images, but also biomedical and aerial images, and to feature point extraction under drastic weather and lighting changes. I will then show that it is also possible to learn efficient object description based on low-level features for scalable 3D object detection.

Conferencia: “Aprendizaje directo y aplicaciones en percepción”

Ponente: Dr. David Jacobs, Universidad de Cincinnati (EEUU) & Universidad Autónoma de Madrid
Viernes, 10 de Abril
Seminario 22, Edificio Ada Byron
Horario: 16:30h-17:30h

Título: Aprendizaje directo y aplicaciones en percepción
En los últimos años se ha desarrollado una aproximación dinámica al estudio del aprendizaje perceptivo y perceptivo-motor. Este marco teórico se denomina teoría del aprendizaje directo. Esta aproximación describe el aprendizaje como un movimiento lento y continuo en un espacio de pocas dimensiones, habitualmente denominado espacio de información-calibración. Cada punto de un espacio de información-calibración representa una tupla de una variable informacional junto a una calibración específica. El aprendizaje es definido, por tanto, como un cambio en la variable informacional que es operativa en un momento dado y un cambio en la forma en que dicha variable es usada. El elemento central de esta teoría es que el aprendizaje mismo -el movimiento en ese espacio- es visto como un proceso guiado informacionalmente. En la presente charla se introduce la teoría de aprendizaje directo y se considera algunas posibles aplicaciones.

Dr. David Jacobs, Universidad de Cincinnati (EEUU) & Universidad Autónoma de Madrid David Jacobs es Licenciado en Matemáticas (Universidad de Granada) y en Psicología Experimental (University of Leiden, the Netherlands). Doctor en Human Movement Sciences (Vrije Universiteit, Amsterdam) ha sido investigador en el L’Institut des Sciences du Mouvement Aix-Marseille Université & CNRS, Francia), en el Human Movement Sciences (Vrije Universiteit, Amsterdam) y en el Center for the Ecological Study of Perception and Action de la University of Connecticut (Cincinnati, EEUU). Actualmente se encuentra como investigador en la Universidad Autónoma de Madrid. Su campo de estudio se centra en la teoría del aprendizaje directo.

Model-based Metacontrol for Autonomy

Fecha: 18/5/2015
Ponente: Carlos Hernández Corbato (GlobalIncubator)
Lugar: Seminario A, Edificio I+D+I - Campus Río Ebro
Hora: 16 a 17

Abstract: There is a current demand for more dependable and autonomous technology. To guarantee mission-level resilience, the limited adaptivity of standard control techniques is not enough, there is a need for actual self-engineering by the system themselves at runtime. For that purpose we propose to leverage the engineering models and biologically inspired cognitive capabilities to build functional metacontrol loops capable of self-awareness and self-adaptation. Our approach goes beyond the customary applications in mobile robotics into the deep waters of autonomous systems engineering. A necessary technology for the machines of the future.

Ponente: Carlos Hernández Corbato es investigador en el Autonomous Systems Laboratory de la Universidad Politécnica de Madrid y CSO en Capital Certainty (capitalcertainty.es) y en Global Incubator (globalincubator.com). Especialista en ingeniería de control y desarrollo de sistemas autónomos, su investigación está centrada en el diseño de tecnologías más autónomas y resilientes, para las cuales incorpora conceptos de ciencias cognitivas como el "self-awareness" con el que pretende generar mecanismos de adaptación en arquitecturas de contro

Charla: : Evolution and Analysis of Empirically-Grounded Models of Brain-Body-Environment Systems

Martes, 24 de Febrero. Seminario 22, Edificio Ada Byron. Horario: 12h-13h.

Dr. Eduardo Izquierdo. School of Informatics and Computing (Universidad de Indiana, Bloomington, EEUU).

Abstract: Even relatively simple animals exhibit a remarkable combination of flexibility and robustness in their behavior. My research aims to address this challenge by constructing and analyzing empirically-grounded models of brain-body-environment systems. In the talk, I will argue that the nematode worm Caenorhabditis elegans is a convenient initial target for such integrated brain-body-environment modeling of a complete animal. My approach uses artificial evolution to explore the space of unknown electrophysiological parameters of the nervous system necessary to generate organism-like behavior.




Más información «


Charla: Fractal Coordination Dynamics in Motor Performance...and Beyond

Miércoles, 25 de Febrero. Seminario 22, Edificio Ada Byron. Horario: 12h-13h.

Dr. Ralf Cox. Faculty of Behavioural Social Sciences (University of Groninger, The Netherlands).

Abstract: It is becoming increasingly evident that the temporal variability in behavior constitutes a rich source of information about its coordinative basis. In this presentation I will emphasize the fractal nature of coordination dynamics underlying human performance in the motor domain and beyond. Attention will be given to techniques for detecting and quantifying the temporal structure of behavior and to the implications of this structure for coordination dynamics.




Más información «


Charla: Functional Organization of Brain Activity for Human Cognition

Viernes, 27 de Febrero. Seminario 22, Edificio Ada Byron. Horario: 12h-13h

Dr. Juan Vidal. Laboratoire de Psychologie et Neurocognition (Université Pierre Mendes-France, Grenoble, France).

Abstract: How does our brain work? How does it explain the way we behave and how we experience our world? These general questions are approached from the point of view of cognitive neuroscience, a research field which merges cognitive psychology and neuroscience. In this talk I will introduce the general concept of selective neuronal communication which lies at the center of functional network interactions in the so-called brainweb. In animal and human neural systems, selective communication is achieved through the mechanism of neuronal oscillatory synchronization. I will review various human and animal intracranial recording studies which focus on neuronal oscillatory synchronization.




Más información «


Seminario: Experiments with the iCub: modelling oculo-motor control and understanding object affordances

Título: Experiments with the iCub: modelling oculo-motor control and understanding object affordances Día: 19/1/2015
Hora: 12:30-13.30
Lugar: Salon de Actos Ada Byron

Ponente: José Santos-Victor - IST, Institute of Systems and Robotics, Lisbon

I will describe some of the experiments we have carried out with the iCub and how they can help us understanding several aspects concerning human cognition. One set of examples will be devoted to the coordination of gaze movements with arm reaching and grasping movements. I will illustrate how data extracted from human subjects can be used to model the interplay between different biologically plausible control loops. In the second part of the talk I will briefly describe how we have used the iCub to learn and explore object affordances and how the aff19/1/2015ordances can be used as a construct for higher-level cognitive skills

Seminario: Vision-controlled Micro Flying Robots

Lugar: Salón de Actos, Edificio Ada Byron
Día: Lunes 15 de Dic
Hora: 12:00
Ponente: Davide Scaramuzza, University of Zurich

Resumen:
In the last two years, we have heard a lot of news about Micro Aerial Vehicles (MAVs), in the form of small quadrotors. Quadrotors have numerous advantages over ground vehicles: they can easily reach environments that no human can access and have more agility and navigational capabilities than ground vehicles. Unfortunately, their dynamics makes them extremely difficult to control and this is particularly true in absence of external positioning systems, such as GPS.. In this talk, I will give an overview of my research activities on visual navigation of MAVs, from slow navigation (using standard frame-based cameras) to agile flight (using event-based cameras).

Robotics and Perception Group of University of Zurich.

Google + Davide Scaramuzza.

Bio:
Davide Scaramuzza is Assistant Professor of Robotics at the University of Zurich and Adjunct Faculty at ETH Zurich. He is founder and director of the Robotics and Perception Group, where he develops cutting-edge research on low-latency vision and visually-guided micro aerial vehicles. He received his PhD (2008) in Robotics and Computer Vision at ETH Zurich (with Roland Siegwart). He was Postdoc at both ETH Zurich and the University of Pennsylvania (with Vijay Kumar and Kostas Daniilidis). For his research contributions, he was awarded an ERC Starting Grant (2014), the IEEE Robotics and Automation Early Career Award (2014), a Google Research Award (2014), the European Young Researcher Award (2012), and the Robotdalen Scientific Award (2009). He is coauthor of the 2nd edition of the book “Introduction to Autonomous Mobile Robots” (MIT Press). He is author of the first open-source Omnidirectional Camera Calibration Toolbox for MATLAB, also used at NASA, Bosch, and Daimler.

Seminario: Duality of Controllability-Observability in a class of Hybrid Systems

Lugar: Seminario del Departamento
Día: 3 miércoles de Diciembre
Hora: 12h
Ponente:José Luis García Malacara.

We are concerned with controllability and observability of timed continuous Petri nets (TCPN) under the infinite server semantics. In particular, we deal with structural and generic controllability (i.e., controllability for all values of the firing rates, and for almost all values of those parameters, respectively). Using ideas developed for linear structured systems theory, the concept of generic controllability is employed to characterize controllability for TCPN in terms of the structure of the Petri net graph. Joins, conflicts, forks and attributions are used to explain how control signals are transmitted or blocked through the net. Based on the net graph, the obtained results are viewed as the dual of the previous obtained for structural and generic observability in TCPN. Clear relationships are stated using the duality concept on Petri net graphs.

Seminario: Bifurcations in Timed Continuous Petri Nets

Lugar: Seminario del Departamento.
Día: 3 miércoles de Diciembre.
Hora: 13h.
Ponente: Manuel Navarro.

The dynamic behavior of TCPN under the infinite server semantics will be considered. In particular, we study the change of the throughput in steady state due to the variation of a firing rate. Jumps in this value (equivalently in the steady state marking) are considered as bifurcations. They occur in nets in which the equilibrium marking of a region goes from being stable to unstable. In _equal-conflict_ nets the bifurcation may occur depending on the firing rates of transitions in conflicts; if transitions are also joins, the steady state marking lies in the boundary between regions.
We show that in consistent and conservative _choice-free_ Petri nets the change in the steady state throughput is continuous. This kind of result can be extended to equal-conflict nets when we vary the firing rates of transitions which are persistent. So, the throughput in steady state is continuous, moreover also monotone for these classes of nets.
Similarly, the variation of the initial marking is considered. There are relationships between the observed behavior of the throughput due to the change of a firing rate and the variation of components of the initial marking. Bifurcations are critical for the performance of the net system, thus very interesting in order to control them

Seminario: "Tecnicas de mapeo y aprendizaje en sistemas agiles (Robots y computadoras vestibles)"

- Ponente:Dr Walterio Mayol Cuevas

- Breve biografía del ponente
Walterio W. Mayol-Cuevas (M’94) received the B.Sc. degree in computer engineering from the National University of Mexico, Mexico City, Mexico, in 1999, and the D.Phil. degree from St. Annes College, University of Oxford, Oxford, U.K., in 2005. He is currently a Senior Lecturer with the Department of Computer Science, University of Bristol, Bristol, U.K., where he is currently involved in research on computer vision and robotics.

- Lugar y Hora
Jueves 20 de Noviembre, 17:00-18:00

- Breve Resumen
Esta platica hace un resumen de trabajo reciente en Bristol sobre aplicaciones y tecnicas de mapeo visual y aprendizaje en sistemas robooticos y de computacion vestible. Nuestro interes se centra en tecnicas de tiempo real que permitan a plataformas agiles y de pocos recursos computacionales, como cuadcopteros y dispositivos como Google Glass, poder tener mejor informacion de su entorno para poder tanto mejorar su control o bien poder proveer informacion relevante en lugares y tiempos oportunos. En particular se discutiran metodos de relocalizacion en tiempo real, de regresion para aprendizaje en cuadcopteros y de agrupamiento para extraccion de informacion relevante en computadoras visuales vestibles.

Seminario: Publish or Perish, Part 1: Why, When, Where, How much?

Ponente: Juan Domingo Tardós, Dept. Informática e Ingeniería de Sistemas,Universidad de Zaragoza
Dia: Jueves, 13 de Noviembre de 2014
Hora:10:00-12:00h
Lugar: Seminario A24 Edificio Ada Byron

Más información «


Seminario: Publish or Perish, Part 2: How?

Ponente: Juan Domingo Tardós, Dept. Informática e Ingeniería de Sistemas,Universidad de Zaragoza
Dia: Jueves, 20 de Noviembre de 2014
Hora:10:00-12:00h
Lugar: Seminario A24 Edificio Ada Byron

Más información «


Seminario:How to give a good research talk

Ponente: Diego Gutiérrez, Dept. Informática e Ingeniería de Sistemas,Universidad de Zaragoza
Dia: Jueves, 11 de Diciembre de 2014
HOra:11:00-12:30h
Lugar: Seminario A24 Edificio Ada Byron

Más información «


Seminario: Manage your research career. Life after PhD

Ponente: Luis Montesano, Dept. Informática e Ingeniería de Sistemas, Universidad de Zaragoza
Dia: Jueves, 04 de Diciembre de 2014
Hora:11:00-12:30h
Lugar: Seminario A24 Edificio Ada Byron
SeminariosTransversales ISI.pdf

Más información «


Seminario: Green Computing meets Big Data

Green Computing meets Big Data Iván Rodero, Rutgers Discovery Informatics Institute, The State University of New Jersey Lunes, 30 de Junio de 2014, 10:30, Seminario DIIS - Edificio Ada Byron

As computational- and data-enabled scientific applications target extreme scales, energy/power-related challenges are becoming dominating concerns. As a result, it is critical to explore emerging architectures such as ongoing many-core processors and deep memory hierarchies as well as applications from an energy perspective, and investigate associated overheads and tradeoffs. For example, energy/power-efficiency has to be addressed in combination with the quality of the solution, performance, resiliency, and other objectives. At the same time, the deluge of big data does not come only from scientific instruments and simulations, but also from sensor networks and other sources such as web crawling and social networks. Being able to transform this data into insights also presents energy/power-related challenges that need to be addressed in the context of different data analytics frameworks.

In this talk, I will explore these issues and will describe selected ongoing research efforts directed towards the scalability and efficiency of the next generation computing systems, including: in-situ data analytics for combustion simulations and co-processing at extreme scales; a framework for exascale co-design; energy/power-efficient data management for big data analytics on systems with deep memory hierarchies and accelerators; and other relevant efforts focused on understanding the cost of data movement and computation scheduling.

Iván Rodero is Assistant Research Professor in the department of the Electrical and Computer Engineering, at the Rutgers Discovery Informatics Institute and at the NSF Cloud and Autonomic Computing Center at Rutgers, The State University of New Jersey. He received an integrated BS and MS degree in Computer Engineering in 2004 and a PhD degree in 2009 from Technical University of Catalonia, Spain. Before joining Rutgers, he was permanent researcher at Barcelona Supercomputing Center. His research interests fall in the broad area of parallel and distributed computing and include high performance computing, data management, energy efficiency, and autonomic and cloud computing. Rodero’s current research activities are in the intersection of green computing and scientific (big) data management and analytics, and have been funded by the National Science Foundation and the U.S. Department of Energy. His research also comprises distributed computing systems, including cloud computing mechanisms and abstractions, federation of large-scale computing systems and cyber-security. He is senior member of IEEE and member of ACM and AAAS.

HTML tutorial



Seminario: Using discrete event system tools for modeling and planning mobile robots

Ponente: Marius Kloetzer (Technical University of Iasi, Romania)

Jueves 8 de mayo de 2014, de 10:00 a 11:00 horas en el seminario del Departamento de Informática e Ingeniería de Sistemas, planta 1ª del Edificio Ada Byron

Abstract. The talk informally presents some techniques for solving problems with immediate applications to mobile robots. The common point of these techniques is to develop a discrete event system that models the environment and/or the robot movements. Based on the mission requirements and on the problem assumptions, we investigate different types of discrete-event models, ranging from automata/graphs to Petri nets. Feasible solutions on the finite state models are found by involving various tools, such as translations to typical optimization problems or resource allocation techniques. Most of the developed frameworks are applicable for a whole team of cooperating robots. The determined discrete solutions yield sequences of simple decisions and routines that should be executed by the mobile robots such that the given task is accomplished.

Speaker's short biography. M. Kloetzer received B.S. (2002) and M.Sc. (2003) degrees in Computer Science from the Technical University of Iasi, Romania, and Ph.D. (2008) in Systems Engineering from Boston University, Boston, MA, USA. He is currently an associate professor at the Technical University of Iasi, Romania. In May-June 2014 he is a guest researcher at the University of Zaragoza, within a mobility grant for guest researches at Campus Iberus. His research interests include planning algorithms for mobile robots based on discrete event systems and high-level specifications.

Seminario: Mecanismos de Mejora del Rendimiento y Calidad de Servicio en Redes Wi-Fi

Impartido por Santiago Pérez, de la Universidad Tecnológica Nacional, Mendoza, Argentina el próximo miércoles 7 de mayo, de 11:00 a 12:00 h., en el Seminario del DIIS (Edificio Ada Byron)

Más información «


Seminario: Realidad Aumentada para Cirugía Laparoscópica

Virtual and Augmented Reality for Laparoscopic Surgery Guidance
Dr. Stéphane André Nicolau, Institut de Recherche contre le cancer de l'appareil digestif, Strasbourg, FRANCE
Jueves 3 de Abril de 2014, 17:00 horas
Seminario A23, Edificio Ada Byron

.

. .

Más información «


Seminario: Color Management in Tone Reproduction

Color Management in Tone Reproduction.
Tania Pouli, Technicolor Research & Innovation.
Fecha: 12 de Diciembre de 2013 (jueves)
Hora: 12:00-13:00h
Lugar: Seminario DIIS - Edificio Ada Byron



Seminario: DeepSee e Iknow de InterSystems: Análisis de la información estructurada y no estructurada en tiempo real

Presenta: Pierre Yves Duquesnoy (Sales Engineer), InterSystems Iberia

Fecha: 28 de noviembre de 2013 (jueves)
Lugar: Aula A12, Edificio Ada Byron.
Hora: 12:00.
Duración prevista: 1h30min (se prevé un descanso de 10min a las 13:00).

Resumen:
-Introducción general a la tecnología de InterSystems.
-Funcionalidades de DeepSee y visión general de la tecnología Iknow.
-Demostración y casos de uso donde se verá cómo gracias a la utilización de DeepSee se nos facilita el análisis en tiempo real basado en información estructurada y no estructurada.

Seminario: Understanding Life through Modelling in an Era of Data Enriched Biology

Duygu Dikicioglu, Department of Biochemistry & Cambridge Systems Biology Centre, University of Cambridge.
Fecha: 28 de Noviembre de 2013 (jueves)
Hora: 11:00-12:00h
Lugar: Seminario DIIS - Edificio Ada Byron

The biotechnology industry has grown in an unprecedented way in recent years owing to advances in fields like molecular modelling, systems biology, disease characterization, or pharmaceutical discovery. Systems biology aims to achieve a quantitative phenotypic description of a biological system. Advances in this particular field have been incorporated in fundamental and applied research. Efforts were focused mainly on the high-throughput analysis of cell parts and genome-scale metabolic modelling. The utilization of such approaches in biotechnology aims to achieve a systems level understanding of the microorganisms that are suitable models in biomedical research or are efficient cell factories for the production of high value commodities. Currently, vast amounts of data are being produced with the new experimental technologies, producing a bottleneck in their analysis. Improved bioinformatics tools and models are currently being developed and employed to capitalize the data on becoming accessible in biotechnology applications. This talk will focus on currently faced challenges in systems biology and several initiatives taken towards addressing several different problems in the topic.

Short-bio:
Duygu Dikicioglu is a research associate in the Department of Biochemistry and Cambridge Systems Biology Centre at Cambridge University. She joined Steve Oliver’s group in Cambridge University in 2012 after the completion of her doctoral degree at Bogazici University, Turkey. She holds PhD, MSc and BSc degrees in Chemical Engineering. Her Bachelor of Science and Master of Science degrees are also earned at Bogazici University. Dr Dikicioglu’s current research focuses on the development and use of systems biology modelling tools to reconstruct biological networks to be used in biotechnology applications from the genomic and post-genomic data.

Más información «


Seminario: Pink Noise Revolution: quantifying political self-organization in social media

Miguel Aguilera, Dept. Infomática e Ing. Sistemas, Universidad de Zaragoza
Viernes, 22 de Noviembre de 2013, 12:00h, Aula A23 - Edificio Ada Byron


Más información «


Seminario: Hacking the NFC cards for fun and honor degrees

Ricardo J. Rodríguez, Technical University of Madrid, Madrid, Spain

Viernes, 15 de Noviembre de 2013, 12:00-13:00, Aula A23 - Edificio Ada Byron

Más información «


Seminario: Virtual and Real World Adaptation for Pedestrian

Antonio López Peña, Computer Vision Center (CVC), Universitat Autònoma de Barcelona (UAB).
Miércoles, 6 de Noviembre de 2013, 12:00-13:00h, Seminario del DIIS - Edificio Ada Byron

Pedestrian detection is of paramount interest for many applications. Most promising detectors rely on classifiers trained with annotated samples. However, the annotation step is a human intensive and subjective task worth to be minimized. By using virtual worlds we can automatically obtain precise and rich annotations. Thus, we face the question: can a pedestrian appearance model learnt in realistic virtual worlds work successfully for pedestrian detection in real world images? Conducted experiments show that virtual-world based training can provide excellent testing accuracy in some real-world datasets, but it also appears the dataset shift problem in many others. Accordingly, during the last years we have explored different domain adaptation ideas for several state-of-the-art pedestrian detection approaches. In this talk, we review this work. Therefore, we will report adaptation results for different features as HOG, LBP, Haar and EOH, different pedestrian models as the Holistic-SVM/AdaBoost and Deformable Part-Based Models (DPM), as well as different strategies as supervised/unsupervised and batch/incremental. Overall, we will see that we have been able to actually adapt virtual and real worlds.

Más información «


Seminario: Publish or Perish, Part 1: Why, When, Where, How much?

El Master en Ingeniería de Sistemas e Informática organiza 4 seminarios sobre temas transversales de investigación, abiertos a todos los interesados:

Publish or Perish, Part 1: Why, When, Where, How much?
Juan D. Tardós, Dept. Informática e Ingeniería de Sistemas, Universidad de Zaragoza
Viernes, 8 de Noviembre de 2013, 10:00-12:00h, Aula A23 - Edificio Ada Byron


Más información «


Seminario: Publish or Perish, Part 2: How?

El Master en Ingeniería de Sistemas e Informática organiza 4 seminarios sobre temas transversales de investigación, abiertos a todos los interesados:

Publish or Perish, Part 2: How?
Juan D. Tardós, Dept. Informática e Ingeniería de Sistemas, Universidad de Zaragoza
Viernes, 15 de Noviembre de 2013, 10:00-12:00h, Aula A23 - Edificio Ada Byron


Más información «


Seminario: How to give a good research talk

El Master en Ingeniería de Sistemas e Informática organiza 4 seminarios sobre temas transversales de investigación, abiertos a todos los interesados:

How to give a good research talk
Diego Gutiérrez, Dept. Informática e Ingeniería de Sistemas, Universidad de Zaragoza
Viernes, 22 de Noviembre de 2013, 10:00-11:30h, Aula A23 - Edificio Ada Byron


Más información «


Seminario: Manage your research career. Life after PhD


Manage your research career. Life after PhD.
Luis Montesano, Dept. Informática e Ingeniería de Sistemas, Universidad de Zaragoza
Viernes, 13 de Diciembre de 2013, 10:00-11:00h, Aula A23 - Edificio Ada Byron

Más información «


Seminario: Fault Diagnosis Graph of Time Petri Nets

Room: Seminario 23. Edificio Ada Byron
Date: 20 of June.
Hour: 12:00

Talk given by Xu Wang.
Abstract: A Discrete Event System (DES) is a dynamic system that evolves in accordance with the abrupt occurrence of events. Faults correspond to discrete events modeling abnormal behaviors and diagnosis is the process to detect faults. In a manufacturing system, a fault may be the failure of a certain operation, e.g., a wrong assembly, or a part put in a wrong buffer. In order to ensure the correct and safe function of large systems, fault diagnosis has attracted a significant attention in DES, untimed and timed systems. The increasing of complexity of systems makes the efficiency of fault diagnosis to be critical. In this work, an online approach for fault diagnosis of timed DES modeled by time Petri net is presented. In the net system, the firings of some transitions are not observable, while others are observable. The occurrence of observable transitions is always observable by an external observer while the occurrence of unobservable ones is not. Faults correspond to a subset of unobservable transitions. In accordance to most of the literature on DES, three diagnosis states, namely normal, faulty and uncertain states, are defined. The proposed approach uses a fault diagnosis graph, which is adapted from state class graph by keeping only the necessary information for computation of the diagnosis states and removing the unnecessary states. Algorithms to compute after each observation only the part of the fault diagnosis graph required to update the diagnosis states are given.

Title: Complexity Analysis of Continuous Petri Nets.
Talk given by Estíbaliz Fraca.
Abstract: Continuous Petri nets were introduced to avoid the state explosion problem of discrete Petri nets. Since then several works have established that the computational complexity of deciding some standard behavioural properties of Petri nets is reduced in this framework. Here we first establish the decidability of additional properties like boundedness and reachability set inclusion. We also design new decision procedures for the reachability and lim-reachability problems with a better computational complexity. Finally we provide lower bounds characterising the exact complexity class of the boundedness, the reachability, the deadlock freeness and the liveness problems.
This work is about continuous Petri nets, but it focus in decidability and complexity analysis, from theoretical computer science.

Seminario: Underwater Robotics Vision

Underwater Robotics Vision
Navid Nourani-Vatani, Australian Centre for Field Robotics.
Viernes, 21 de junio de 2013, 10:00-11:00, Seminario DIIS, Edificio Ada Byron

Abstract
In this talk I'll be giving an overview of the underwater robotics and computer vision work that we carry out in the marine robotics group at ACFR. I'll introduce our robots briefly and then dig into the challenges we are faced operating a robot out of reach in depths of up to 250m. I'll then discuss the difficulties of using vision under water and our approaches to color correction, 3D stitching, clustering and classification as well as show some results.

Navid Nourani-Vatani received a B.E.E. and a M.Sc. in electrical and electronics engineering from the Technical University of Denmark (DTU) in 2004 and 2007, respectively. Between 2006-2011 he was with the Autonomous Systems Laboratory at the CSIRO ICT Centre developing coverage, localization, and navigation technologies for autonomous ground vehicles. In 2011 he received his PhD from University of Queensland engaged with the research topic of biologically-inspired vision-based localization algorithms. He is currently at the Australian Centre for Field Robotics developing image-based benthic habitat classification and counting for Autonomous Underwater Vehicles..

Seminario válido para la asignatura 62621 "Seminario de Linea de Investigación"
(2 ECTS) del Master en Ingeniería de Sistemas e Informática.
Las transparencias de los seminarios estarán disponibles en:
https://moodle.unizar.es/course/view.php?id=2201

Seminario: Action Switching in Brain-Body-Environment Systems

Action Switching in Brain-Body-Environment Systems
Eran Agmon, Cognitive Science and Complex System Dpt., Indiana University.
Viernes, 31 de Mayo de 2013, 12:00, Aula A23 - Edificio Ada Byron

Abstract
Research on dynamical systems in the cognitive and behavioral sciences studies how systems made of many coupled components across brain, body, and environment, self-organize to generate behavior. This approach has mostly focused on modeling single actions and has not addressed how a single dynamical system can engage in multiple different directed actions. In this talk, we will examine computational models of dynamical agents that engage in multiple actions and autonomously switch between them purely through self-organized dynamics. The analysis of one of these agents shows how different actions can arise through transient modes of sensorimotor coordination, in which a subset of the available sensors and effectors become engaged and others are ignored by making use of sensorimotor invariants and neuronal gating mechanisms. Transitions between actions are analyzed and shown to be rapid movements through the agent’s neuronal state space. By undergoing these transitions, some of the previously controlling sensors and effectors disengage, and new sets of sensors and effectors are engaged.

Eran Agmon works at the NSF IGERT for the dynamics of brain-body-environment systems within the Cognitive Science Program at Indiana University. His research is focused on using the framework of complex systems science to study cognition as a dynamical, embodied, and situated process that unfolds through the interactions between the brain, body, and environment. He received his BS in Cognitive Science at the University of California, San Diego and his MS in Systems Science at Portland State University..

Seminario válido para la asignatura 62621 "Seminario de Linea de Investigación"
(2 ECTS) del Master en Ingeniería de Sistemas e Informática.
Las transparencias de los seminarios estarán disponibles en:
https://moodle.unizar.es/course/view.php?id=2201

Seminario: Visual Navigation and 3D Reconstruction for Flying Robots

Visual Navigation and 3D Reconstruction for Flying Robots
Jürgen Sturm, Technical University of Munich.
Jueves 23 de Mayo de 2013, 11:00, Seminario DIIS, Edif. Ada Byron

Abstract
Autonomous navigation with quadrocopters in unknown 3D environments requires accurate and robust solutions for localization and mapping. In my talk, I will present recent approaches for monocular and RGB-D cameras, and how they can be applied to quadrocopters. In particular, I will focus on dense methods for visual odometry and 3D reconstruction that do not require the extraction of visual features.

Bio: Dr. Jürgen Sturm is a Post-Doc in the Computer Vision and Pattern Recognition group of Prof. Daniel Cremers at the Department of Computer Science of the Technical University of Munich. His major research interests lie in dense methods for localization and mapping for micro aerial vehicles. In 2011, he obtained his PhD from the Autonomous Intelligent Systems lab headed by Prof. Wolfram Burgard at the University of Freiburg. He received the ECCAI Artificial Intelligence Dissertation Award 2011 for his PhD thesis and was distinguished with the TUM Teach Inf Award 2012 for his course "Visual Navigation for Flying Robots".

Seminario válido para la asignatura 62621 "Seminario de Linea de Investigación"
(2 ECTS) del Master en Ingeniería de Sistemas e Informática.
Las transparencias de los seminarios estarán disponibles en:
https://moodle.unizar.es/course/view.php?id=2201

Seminario: Semantic Parsing for Priming Object Detection in RGB-D Scenes

Semantic Parsing for Priming Object Detection in RGB-D Scenes
César Darío Cadena Lerma, George Mason University
Jueves 16 de Mayo de 2013, 11:00, Seminario DIIS, Edif. Ada Byron

Abstract:
The advancements in robot autonomy and capabilities for carrying out more complex tasks in unstructured indoors environments can be greatly enhanced by endowing existing environment models with semantic information.
In this talk we describe an approach for semantic parsing of indoors environments into semantic categories of Ground, Structure, Furniture and Props.
Instead of striving to categorize all object classes and instances encountered in the environment, this choice of semantic labels separates clearly objects and non-object categories.
We use RGB-D images of indoors environments and formulate the problem of semantic segmentation in the Conditional Random Fields Framework. The appearance and depth information enables us induce the graph structure of the random field, which can be effectively approximated by a tree, and to design robust geometric features, which are informative for separation and characterization of different categories. These two choices notably improve the efficiency and performance of the semantic parsing tasks.
We carry out the experiments on a NYU V2 dataset and achieve superior or comparable performance and the fraction of computational cost than the state of the art.

César Darío Cadena Lerma is currently a Post-doctoral Research Fellow at CS Department, George Mason University. He got his PhD, under the supervision of José Neira, in the Department of Computer Science and System Engineering at the University of Zaragoza. His main research interest is in robotics and control, in which he is particulary focused in Simultaneous Localization and Mapping problems, persistent mapping in dynamic environments, as well as the use of machine learning techniques for Semantic Data Association and Place Recognition tasks. http://www.cs.gmu.edu/~ccadenal/
----------------------------------------------------------------------------------
Seminario válido para la asignatura 62621 "Seminario de Linea de Investigación" (2 ECTS) del Master en Ingeniería de Sistemas e Informática. Las transparencias de los seminarios estarán disponibles en: https://moodle.unizar.es/course/view.php?id=2201

Seminario: Compressive Light Field Displays

Gordon Wetzstein, MIT Media Lab

Viernes 3 de Marzo de 2013, Seminario Dpto, Edificio Ada Byron, 1100.

The core goal of my research is the design and implementation of novel approaches that push the boundaries of conventional light capture and display technology through the co-design of optics and computational processing. Such computational imaging systems facilitate new capabilities such as high-resolution, glasses-free 3D displays. The talk will focus on three recently developed technologies:

1-Layered 3D displays: A compressive light field display using stacked, inkjet-printed transparencies in combination with computed tomographic light field synthesis. This design allows for the fabrication of inexpensive, high-resolution, and bright light field displays showing static content.

2-Polarization Fields: An optically and computationally efficient compressive light field design. Multiple stacked LCDs, stripped off their polarizing films and color filters, act as programmable polarization rotators. Combined with efficient solvers implemented on the GPU, real-time framerates for dynamic content are achieved.

3-Tensor Displays: A new family of compressive light field displays comprising all architectures employing stacks of high-speed LCDs illuminated by uniform or directional backlighting (i.e., any low-resolution light field emitter). Supports wider fields of view. larger depths of field, and thinner form factors than previous layered displays.

Charla de introducción a metodologías ágiles de desarrollo de software

Se ha organizado para la semana que viene una charla de introducción a metodologías ágiles de desarrollo de software. Es de particular interés para aquellos que están trabajando en pequeños grupos de desarrollo de I+D y profesores relacionados con asignaturas de Ingeniería del Software. La entrada es libre.:
¿Te suena Scrum y Extremme programing? ¿Sabes lo que esel Kanban? :
¿Has oído hablar de las metodologías ágiles?:
Si no has oído hablar de ellas es mejor que empieces, porque va a ser la forma más habitual de gestionar proyectos y equipos dentro de unos años.:
Al menos en parte del mundo ya lo es.:
Desde el grupo Agile Aragón se ha organizado un evento introductorio a las metodologías ágiles, para ayudarte a resolver todas esas dudas y muchas más.:
Será el martes 6 de Noviembre a las 19h en el Aula A.11 de la EINA (Edificio A) y se tratarán los siguientes temas en 4 pequeñas charlas::
• Agile Manifesto, por Dani Latorre (@dani_latorre):
• Scrum, por Fernando Pérez (@ferpega_):
• Extreme Programming, por Julio García (@papajulio):
• Kanban, por Teresa Oliver (@tolivern):


Modeling and Verification of Hybrid Systems

Recent technological innovations have caused a considerable interest in the study of dynamical processes of a mixed continuous and discrete nature. Such processes are called hybrid systems and are characterized by the interaction of time-continuous models (governed by differential or difference equations) on the one hand, and logic rules and discrete-event systems (described by, e.g., automata, finite state machines, etc.) on the other. In practise a hybrid system arises, e.g., when continuous physical processes are controlled via embedded software that intrinsically has a finite number of states.
An overview of the field of hybrid systems will be offered first. Then aspects related to modelling and verification through reachability methods and abstractions will be particularly emphasised.*

Prof. JANAN ZAYTOON
Professor of Automatic Control, Université de Reims Champagne-Ardenne, his main research area is Modelling, simulation, analysis, validation and control of discrete event and hybrid systems.
Director of CReSTIC (Centre de Recherche en Science et Techniques de l’Information et la Communication de l’Université de Reims), a center that groups 150 members (25 Professors, 60 Lecturers and Associate Professors, 50 PhD students, 10 Post-doc, 10 technical and administrative persons) and Director of GDR MACS (Groupement du Recherche en Modélisation, Analyse et Commande des Systèmes réactifs) of CNRS, a national research network in charge of the scientific acivity of the French Community in Automatic Control and Production Systems (about 2.000 researchers involved).
IFAC Council Member and Head of the IFAC French NMO (National Member Organizer), he has published more than 200 journal & conference papers, books, book chapters and patents. Editor of /Control Engineering Practice/ and /Nonlinear Analysis: Hybrid Systems/; Associate Editor of /Discrete Event Dynamic Systems/, /IET Control Theory and Applications/, and /International Review of Automatic Control/; is Member of the Publication Committee of IFAC, and editor of 20 special journal issues and 16 conferences.

TRES SESIONES (día, hora y lugar):
1. martes 10/02 de 16:00 a 19:00 (Seminario del Depto)
2. miércoles 11/02 de 11:00 a 14:00 (Seminario24, 2ªplanta)
3. jueves 12/02 de 16:00 a 19:00 (Seminario del Depto)


Aprendizaje a partir de flujos de datos: Teoría y Aplicaciones

Seminario impartido por el profesor Ricard Gavaldà, de la UPC, sobre el modelo de aprendizaje Data Stream, del que es un experto reconocido.

Fechas:
- miércoles 29 y jueves 30 de abril (antes del puente del 1 de mayo)
- quince días después: jueves 14 y viernes 15 de mayo

PROGRAMA RESUMIDO:

1. Modelos de flujos de datos y algorítmica
1.2. Flujos de datos: motivación
1.3. Recordatorio: estadística y probabilidad
1.4. Modelos de flujos de datos
1.5. Estimación de momento y entropía
1.6. Actualización de histogramas
1.7. Cotas inferiores
1.8. Procesamiento de consultas y transacciones
1.9. Detección y cuantificación de cambio de concepto

2. Aprendiendo a partir de flujos de datos
2.2. Contextualización del aprendizaje automático y de la minería de datos
2.3. Clasificación
2.4. Agrupamientos
2.5. Descubriendo ítems frecuentes
2.6. Ventanas adaptativas

3. Ejemplos de aplicación: monitorizando tráfico en Internet
3.2. Temas no cubiertos en el curso
3.3. Líneas de investigación actuales


Animación interactiva de contactos y deformaciones

Ponente: Miguel Angel Otaduy, Univ. Rey Juan Carlos
Hora: Lunes 1 de Diciembre a las 16:00h
Lugar: Seminario A13, Edificio Ada Byron

Abstract:
En el mundo que nos rodea, las leyes físicas hacen que los objetos choquen unos contra otros, se deformen, se rompan, y den lugar a efectos y movimientos de gran riqueza y difíciles de clasificar. El realismo de las aplicaciones de realidad virtual, animación y videojuegos depende en gran medida de la fidelidad con que se simulen los efectos de deformación y contacto entre objetos, pero la precisión de los algoritmos choca con las necesidades de eficiencia de las aplicaciones interactivas. Esta conferencia incluirá una breve descripción de los problemas técnicos presentes en la simulación de contacto y deformaciones, así como la exposición visual de distintos avances recientes.

Seminario válido para la asignatura optativa
62621 "Seminario de Linea de Investigación" (2 ECTS)

Aplicaciones de la Programación Entera en el área de Arquitectura de Computadores

Ponentes: Clemente Rodríguez (U. País Vasco) y Luis Carlos Aparicio (UZ, Teruel)

Fecha: Jueves 10 y Viernes 11, de 10h a 13h

Lugar: Seminario del departamento

Contenido:
1.- Soporte a los compiladores/optimizadores:
- Optimal Instruction Scheduling
- Dependence Analysis

2.- Introducción al análisis del Worst-Case Execution Time
- Métodos
- Herramientas

3.- Análisis del Worst-Case Execution Time de un programa
- Modelado de la Memoria Cache:
-Contenido: Instrucciones, datos
-Asociatividad
-Algoritmo de reemplazo
-Cache Miss Equations
-Anotación de falsos caminos
-Modelado del predictor de saltos y ejecución especulativa
-Modelo de ejeción del procesador: simple issue, multiple issue (in-order...)

4.- Worst-Case Energy Consumption



Seminario Investigación. Interoperabilidad de Sistemas de Información en Internet

Título: "Issues concerned with the development of a European Spatial Data Infrastructure" Ponente: Dr. Ioannis Kanellopoulos, Spatial Data Infrastructures Unit, Institute of Environment and Sustainability, Joint Research Centre, European Commission (Italy) Fecha: Lunes 26 de marzo de 2007 Hora: 9:30 h. Lugar: Seminario 25 del Edificio Ada Byron Duración: 60 min Resumen: La presentación tratará sobre el estado actual de las Infraestructuras de Datos Espaciales en Europa y en especial sobre las implicaciones tecnológicas derivadas de la aprobación de la Directiva Europea INSPIRE (Infrastructure for Spatial Information in Europe). El Dr. Ioannis Kanellopoulos describirá además los objetivos y areas de investigación abordadas actualmente dentro del Instituto de Medio Ambiente y Sostenibilidad del Centro de Investigación Conjunto (JRC) de la Comisión Europea.

Seminario de Investigación del DIIS

Título: "Software Demand, Hardware Supply"
Ponente: Jesús Alastruey Benedé (Profesor DIIS, Investigador GaZ)
Viernes 15 de diciembre de 2006, 12:00h, SemDIIS


Seminario de investigación del DIIS

"On controllability in timed continuous Petri net systems",
Renato Vázquez (Becario del GISED, DIIS),
Viernes 1 de diciembre de 2006, 11:00h, SemDIIS.

Más información «


Seminario de investigación del DIIS

"Tailoring Robot Actions to Task Contexts using Action
Prediction Models", PhD Freek Stulp (Intelligent Autonomous
Systems Group, TUM, Munich, Germany),
2 de noviembre de 2006, 12:00h, SemDIIS.

Más información «


[26-06-06] Dr. J. M. M. Montiel (Profesor Titular del DIIS) y J. Civera (Profesor Ayudante del DIIS)

SLAM con cámara monocular.

[02-06-06] Dr. D. Josep Solé i Pareta (Catedrático de Arquitectura y Tecnología de Computadores, UPC)

Características de funcionamiento de las redes de conmutación de paquetes ópticos.

[15-05-06] D. Antonio Ruiz-Cortés (Universidad de Sevilla)

Improving the automatic procurement of web services using constraint programming.

[18-04-06] Dr. Ian Reid (University Lecturer in Engineering Science and Fellow of Exeter College University of Oxford, UK)

Real-time Visual SLAM.

[11-04-06] Dr. Ouri Wolfson (Richard and Loan Hill Professor, Computer Science Department at University of Illinois, Chicago)

A new PhD program in Computational Transportation Science.

[07-04-06] Dr. Jack Lutz (Professor of Computer Science at Iowa State University, USA)

New dimensions in the Theory of Computing.

[05-04-06] Dr. Robyn Lutz (Professor at Iowa State University and Senior Engineer at Jet Propulsion Laboratory, NASA)

Identifying Contingency Requirements using Obstacle Analysis.

[20-03-06] Dr. Xu Jing (Visitante Postdoctoral)

Tracking Control of Join-Free Timed Continuous Petri Net Systems.

[08-03-06] Dr. Carlos Sagüés (Profesor Titular del DIIS)

Localization with Omnidirectional Images using the Radial Trifocal Tensor.

[02-03-06] John. M. Hitchcock (University of Wyoming, USA)

Online Learning and Resource-Bounded Dimension: Winnow Yields New Lower Bounds for Hard S.

[01-03-06] Alejandro Mosteo (PhD Student Robotics, DIIS)

Planificación y asignación de tareas en equipos multi-robot.

[22-02-06] Claudine Chaouiya Chantegrel (Ecole Superieure d’Ingenieurs de Luminy, Marseille, France)

Modelling of biological networks.

[08-02-06] Gonzalo López (PhD Student Robotics, DIIS)

Matriz Fundamental y Visual Servoing.

[19-01-06] Javier Civera (PhD Student Robotics, DIIS)

Homografías y paralaje.

[16-12-05] Dr. Steven LaValle (Department of Computer Science, Univ. Illinois)

Information Spaces: They’re Everywhere.

[13-12-05] Dr. Philippe Moser (Visitante Postdoctoral)

Recent results in derandomization.

[02-12-05] María Villarroya Gaudó (Profesor Ayudante del DIIS)

Sistemas micro/nano-electromecánicos: Como la nanotecnología mejora las prestaciones de muchos dispositivos.

[23-11-05] Lina María Paz (PhD Student Robotics, DIIS)

Single Value Decomposition (SVD).

[16-11-05] Rubén Martínez-Cantín, (PhD Student Robotics, DIIS)

Filtros Gaussianos: EKF y UKF.

[09-11-05] Pedro Piniés and Luis Montesano (PhD Students Robotics, DIIS)

Estimación de parámetros: Efecto de la linealidad.

[28-10-05] Iñaki Raño (Profesor Ayudante Doctor del DIIS)

Evitación de obstáculos usando agresión de Braitenberg y fusión motríz.