Unstructured Video Based Rendering (VBR)
We present an algorithm designed for navigating around a performance that was filmed as a "casual" multi-view video collection: real-world footage captured on hand held cameras by a few audience members. The objective is to easily navigate in 3D, generating a video-based rendering (VBR) of a performance filmed with widely separated cameras. Casually filmed events are especially challenging because they yield footage with complicated backgrounds and camera motion. Such challenging conditions preclude the use of most algorithms that depend on correlation-based stereo or 3D shape-from-silhouettes.
Datasets: | [link] |
Binaries: | [link] |
Publications:
Unstructured Video-Based Rendering: Interactive Exploration of Casually Captured Videos(Oral)SIGGRAPH 2010, Los Angeles, USA |
CrowdCam: Instantaneous Navigation of Crowd Images using Angled GraphA. Arpa, L. Ballan, R. Sukthankar, G. Taubin, M. Pollefeys, R. Raskar [PDF] [video] [bibtex]3DV 2013, Seattle |
Modeling Dynamic Scenes Recorded with Freely Moving CamerasA. Taneja, L. Ballan and M. Pollefeys [PDF] [web] [bibtex]ACCV 2010, Queenstown, New Zealand |
3D Reconstruction and Video-Based Rendering of Casually Captured VideosA. Taneja, L. Ballan, J. Puwein, G. J. Brostow and M. Pollefeys [bibtex]Video Processing and Computational Video [Book]Daniel Cremers, Marcus A. Magnor, Martin R. Oswald, Lihi Zelnik-Manor, Springer, 2011 |