Fourth CVC Workshop on the Progress of Research & Development CVCRD 2009 "New Trends and Challenges in Computer Vision" Computer Vision Center, UAB October 30, 2009 MANUSCRIPT SUBMISSION
General Chairs Xavier Baró Sergio Escalera Miquel Ferrer Organization Committee Josep Lladós Montse Culleré Helena Piulachs Program Committee Jaume Amores Andrew Bagdanov Ramon Baldrich Pau Baiget Simone Balocco Robert Benavente Alicia Fornés Carlo Gatta Débora Gil Jordi Gonzàlez Aura Hernández Laura Igual Carme Julià Dimosthenis Karatzas Ágata Lapedriza Josep Lladós Antonio López Felipe Lumbreras Enric Martí David Masip Mikhail Mozerov Xavier Otazu Carlos Alejandro Párraga Daniel Ponsa Oriol Pujol Petia Radeva Bogdan Raducanu Xavier Roca David Rotger Marçal Rusiñol Anna Sabaté Gemma Sánchez Xavier Sánchez Ángel Sappa Joan Serrat Ricardo Toledo Ernest Valveny Joost Van de Weijer Maria Vanrell Fernando Vilariño Juan José Villanueva Jordi Vitrià Contact Information CVC Research Secretariat Computer Vision Center Edifici O, campus UAB Bellaterra 08193 Catalonia Spain	We are pleased to announce that the Fourth CVC Workshop on the Progress of Research and Development (CVCRD'2009) will be held at the Computer Vision Center, on October 30, 2009. The CVC workshops provide an excellent opportunity for young researchers and project engineers to share new ideas and knowledge about the progress of their work. It sketches the state of the research and development activities in the last year. In addition, the workshop is the welcome event for new people that recently have joined the institute.
	Presenters The submission of manuscripts is open to: It is mandatory for each PhD student. It is good practice in their training process. For some of them it will be as a dry-run of conference participation. Development engineer teams and spin-off companies are also encouraged to submit papers describing projects
	Workshop Format The program will be organized in a single-track half-day workshop. It will comprise several oral sessions dedicated to specific topics. For each session, the chairs will introduce the state art and state the open questions for the session's topic, and it will be followed by a number of short presentations addressed by paper authors. Each session will be concluded by a panel discussion. A poster session will take place during the coffee break.
	Topics The topics of interest cover, but are not limited to, the CVC research lines and development projects at large: Medical Imaging. Robot Vision. Color & Texture Analysis. Object Recognition. Motion Analysis and Recognition. Advanced Driver Assistance Systems. Machine Vision. Document Analysis. Pattern Recognition. Applications.
	Instructions for oral presentations Each oral presentation is scheduled for 10 minutes. At the end of each session it will be a 10-minute discussion time for questions and comments about all the talks in the session. Powerpoint or PDF slides are recommended for your talk. The room will be equipped with a presentation notebook with MS Office, Acrobat Reader, and popular programs like Windows Media Player for playing videos. Please contact us if you need any particular software for this purpose. The presenters are kindly requested to copy their files in the shared folder \\CVC155\CVCRD09. We ask the speakers to create a folder named “SESSION_NumSession_SpeakerSurname”. It is of utmost importance that authors keep to the presentation time limit to achieve appropriate timing. Therefore, please organize your slides around the points you intend to make, use no more than one slide per minute. A reasonable strategy is to allocate about 2 minutes per slide when there are equations or important key points to make, and one minute per slide when the content is less complex. Slides attract and hold attention and reinforce what you say - provided you keep them simple and easy to read Instructions for posters Poster sessions are a good medium for authors to present papers and meet with interested attendees for in-depth technical discussions. In addition, attendees find the poster sessions a good way to sample many papers in parallel. Thus, it is important that you display your message clearly and noticeably to attract people who might have an interest in your paper. Your poster should cover the key points of your work. It need not, and should not, attempt to include all the details, you can describe them in person to people who are interested. The ideal poster is designed to attract attention, provide a brief overview of your work, and initiate discussion. The suggested poster size is A0: 841mm (width) x 1189mm (height). However, we strongly encourage the authors to re-use posters recently used in other conferences that may have very related content. The posters should be mounted during the day before the workshop. The poster session will be held in the inner courtyard. A slot for each author will be allocated in the poster panels. Tape will be provided to mount posters on the boards. Please, note that this year, the organization has introduced a HIGHLIGHT POSTER SESSION in the workshop program. This means that every person who presents a poster must prepare ONE SINGLE SLIDE in order to present his/her poster in that session, following the template given by the organization. The maximum time allowed to present this slide is 45 seconds per person. Highlight Poster Template
	Workshop Program OPENING (9:30 – 9:45) Welcome and Opening Remarks, Josep Lladós 9:45 -10:25 SESSION 1 (ADAS) Chairs: A. Sappa & J. Amores Adaptive Model-based Road Detection using Shadowless Features. J. M. Álvarez and A. M. López Performance of classical monocular egomotion methods in the ADAS context. D. Cheda, D. Ponsa and A. López 10:25 - 11:05 SESSION 2 (Color and Texture) Chairs: A. Párraga & J. van de Weijer Hybrid Fusion: Beyond Early and Late Fusion for Texture Classification. F. Shahbaz Khan, J. van de Weijer, and M. Vanrell Towards non-supervised segmentation: a comparison of goodness measures based on saliency and contrast. E. Vazquez and R. Baldrich Computational Color: Representation, Constancy and psychophysics. J. Vazquez-Corral and M. Vanrell 11:05 - 11:30 POSTER HIGHLIGHT 11:30 - 12:15 COFFEE BREAK & POSTER SESSION Synthetic Urban Development to Evaluate Pedestrian Detection. J. Marín and A. López Feature matching with graphical models for night vehicle detection. J. Carlos Rubio Ballester and J. Serrat Detecting small pedestrians. D. Vázquez, D. Gerónimo, A. López Object Color Alteration. Shida Beigpour and J. van de Weijer Human and Computational Color Constancy. J. Roca, C.A Párraga and M. Vanrell Optimal Feature Selection for Subspace Image Matching. Gemma Roig, Xavier Boix, Fernando de la Torre A Computational Colour Naming Model Trained on Real-Life Images. Hany M. SalahEldeen, R. Benavente, M. Vanrell Perceptual Feature Detection. N. Murray, X. Otazu, M. Vanrell On Experimental Evaluation of Descriptors for Facial Feature Point Detection. M. A. Rojas Q., D. Masip and J. Vitrià Using Colour Saliency for Image Retrieval. J. I. Toledo, J. van de Weijer Text Segmentation in Colour Poster from the Spanish Civil War Era. A. Clavelli, D. Karatzas A rotation invariant page layout descriptor for document classification and retrieval. A. Gordo, E. Valveny Perceptual Criteria on JPEG2000 Quantization. Jaime Moreno, X. Otazu and M. Vanrell Use of Filtered Back-projection Methods to Improve CT Image Reconstruction. J. Bernal and J. Sánchez Towards Detection of Measurable Contractions using WCE. M. Drozdzal, P. Radeva, S. Seguí, F. Vilariño, C. Malagelada, F. Azpiroz and J. Vitrià Adaptive Dynamic Space Time Warping for Real Time Sign Language Recognition. S. Escalera, A. Escudero, P. Radeva, and J. Vitrià An Analysis of Theoretical and Practical Aspects of Spatio-Temporal Regular Flow. J. D. Gomez, C. Gatta, P. Radeva 3D Human Action Recognition using Key Poses, Wenjuan Gong, A. D. Bagdanov and J. Gonzàlez Advances in Variational Optical Flow, N. Onkarappa and A.D. Sappa Image Description using Local Binary Patterns: Application to Scene Classification. N. Elfiky, J. Gonzàlez Graph-based representations for Object Recognition. J. Gibert and E. Valveny Ranking Error-Correcting Output Codes for Class Retrieval. M. Mirza-Mohammadi, F. Ciompi, S. Escalera, O. Pujol, P. Radeva Colour Logo Recognition. F. Nourbakhsh, D. Karatzas and E. Valveny Calibration and Rectification of Multimodal Stereo Rigs. F. Barrera, F. Lumbreras, and A. D. Sappa Quadric Surface Fitting: Orthogonal versus Estimated Distances. M. Rouhani and A. D. Sappa 12:15 - 12:55 SESSION 3 (Document Analysis) Chairs: D. Karatzas & M. Russiñol Coloring Laplacian-of-Gaussian Detector for Image Matching. D. A. Rojas Vigo and J. van de Weijer Comparison of Seal Detection by Different Character Shape Features. P. Pratim Roy, U. Pal and J. Lladós 12:55 – 13:35 SESSION 4 (Motion and Tracking) Chairs: A. Bagdanov & P. Baiget Reactive object tracking with single uncalibrated PTZ camera. M. al Haj, A. D. Bagdanov and J. Gonzàlez Robust Background Subtraction Approach based on Chromaticity and Intensity Patterns. A. Amato, M. Mozerov, and J.Gonzàlez Interest Point based Human Action Recognition. B. Chakraborty, A. D. Bagdanov, and J. Gonzàlez 13:35 – 14:15 SESSION 5 (Descriptors) Chairs: C. Gatta & A. Lapedriza Semantic Segmentation of Images Using Random Ferns. J. Mª Gonfaus, J. Gonzàlez, Theo Gevers Object Detection using Coarse-to-Fine relocalization. M. Pedersoli, J. Gonzalez, A. Bagdanov and J. J. Villanueva BeaStreamer-v0.1: a new platform for Multi-Sensors Data Acquisition in Wearable Computing Applications. P. Casale, O. Pujol, P. Radeva LUNCH (14:15 – 16:00)