Smart Multicamera System - Demos

Tomas Svoboda, Petr Doubek, Karel Zimmermann

Multiple simple cameras in a room. [HW | SW]	What do the cameras see? [video]
Volumetric reconstruction [video]	Event recognition with virtual editor [video]
3D articulated human model [video]	3D tracking [video1] 3D tracking [video2]
SMC (particle filtering) tracker [ video1 | Principle: video2 ]	Fully automatic self-calibration [ video | SW | presentation (short) PDF
Principle: PDF | Demo	Learnable sequential linear predictors. Demo with publications, videos, etc. The tracker was used in the Curling simulator.

Student Demos

• Lukas Fajt. Pictorial Structural Models, Learning and Recognition in Image Sequences
• Ondrej Mazany. Articulated 3D Human Model and its Animation.
• Jan Kosata. Automatic Segmentation of Articulated Objects of Unknown Complexity from Videos.
• Petr Lhotsky. Sequential monte carlo for head detection and tracking. Talk and videos. Thesis (PDF)
• Curling simulator. A team project in summer term 2006/2007. Presentation (in Czech) [PDF]. Playing it: overview video, machine output video.
• Tomas Losert. Cut Placement in a Multi-Camera Sequence.
• Jan Konecny. Adaptive linear predictor suitable for objects with variable appearance.
• Rostislav Prikner. Pictorial Structural Models for Human Detection in Videos. 06/2008
• Ctibor Varga. Object detection by randomly initialized tracking. 06/2008, in Czech.