CVBASE '06 Final program

May 12, Graz University of Technology


9:00 Welcome address

9:15 Invited talk

  • Pascal Fua: Recovering 3D Shape and Motion from Video

Video-based approaches are an extremely cost effective way to analyze athletic performance. However, full automation remains elusive because image-data is often incomplete, imprecise or, even, full of errors. In our work, we have developed techniques that rely on sophisticated 3D shape and motion models to fit such noisy and incomplete data.

In this talk, I will present these models and the statistical learning techniques we use to learn them from training data. I will also demonstrate their effectiveness for applications, such as those depicted by the figure below, which range from analyzing a golf swing to recovering the 3D shape of a spinnaker.

10:00 Session: Player/athlete tracking

  • Wei Du, Jean-Bernard Hayet, Justus Piater, and Jacques Verly: Collaborative Multi-Camera Tracking of Athletes in Team Sports

  • Matej Kristan, Janez Pers, Matej Perse, and Stanislav Kovacic: Towards fast and efficient methods for tracking players in sports

11.00 Coffee break

11.15 Session: Ball tracking

  • Takumi Shimawaki, Jun Miura, Takuro Sakiyama, and Yoshiaki Shirai: Ball Route Estimation in Broadcast Soccer Video

  • Jinchang Ren, James Orwell, and Graeme A. Jones: Generating Ball Trajectory in Soccer Video Sequences

12.15 Lunch

14.15 Session: Motion analysis and activity recognition 1

  • Wei-Lwun Lu and James J. Little: Tracking and Recognizing Actions at a Distance

  • Scott J. Blunsden, Robert B. Fisher, and Ernesto Andrade: Recognition of Coordinated Multi Agent Activities, The Individual vs The Group

15.15 Poster session and coffee break

  • Xavier Desurmont, Jean-Bernard Hayet, Jean-Francois Delaigle, Justus Piater, and Benoit Macq: Trictrac Video Dataset: Public HDTV Synthetic Soccer video sequences with Ground truth.

  • Javier Ruiz-del-Solar, Patricio Loncomilla, Paul Vallejos, and Pablo Gonzalez: An Automated Refereeing and Analysis Tool for Robot Soccer

  • Vincenzo Caglioti and Alessandro Giusti: Ball trajectory reconstruction from a single long-exposure perspective image

16.00 Session: Motion analysis and activity recognition 2

  • Matej Perse, Matej Kristan, Janez Pers, and Stanislav Kovacic: A Template-Based Multi-Player Action Recognition of the Basketball Game

  • Ryan Cassel: Acrobatic Movement Analysis Based on Computer Vision

17.00 Close



















ELATED PageKits © 2003 ELATED.com/PageKits.com