|
|
CVPR 2009 Video Review
The CVPR 2009 Video Review is an opportunity to show off the newest
techniques and results from our community. Many algorithms are best
explained visually as animations. Some challenging problems and results
can only really be appreciated as videos. Systems are getting better and
faster, and deserve to be seen by a broader audience. If your results are
visually interesting, then this is your chance to be recognized, and to be
seen both during the conference, and afterward, when attendees get back to
their classrooms, companies, and labs. The compilation will play
prominently in a loop on display at the main conference, and will be
included in the CVPR 2009 Proceedings DVD-ROM.
New! Accepted videos announced here!
Video
Title |
Author
Names |
|
|
Markerless Motion Capture with
Unsynchronized Moving Cameras |
Nils
Hasler*, MPI Informatik
Bodo Rosenhahn, Hannover University
Thorsten Thormaehlen, MPI Informatik
Michael Wand, Saarland University
Juergen Gall, BIWI, ETH Zurich
Hans-Peter Seidel, MPI Informatik |
Predicting High Resolution Image
Edges with a Generic, Adaptive, 3-D Vehicle Model |
Matthew
Leotta*, Brown University |
Motion Capture Using Joint
Skeleton Tracking and Surface Estimation |
Juergen
Gall*, BIWI, ETH Zurich
Carsten Stoll,
Edilson De Aguiar,
Christian Theobalt,
Bodo Rosenhahn, Hannover University
Hans-Peter Seidel, MPI Informatik |
Rank Priors for Continuous
Non-Linear Dimensionality Reduction |
Andreas
Geiger*, KIT
Raquel Urtasun, EECS Berkeley
Trevor Darrell, EECS Berkeley |
Similarity Metrics and Efficient
Optimization for Simultaneous Registration |
Christian
Wachinger*, TU München
Nassir Navab, TU München |
LidarBoost: Depth Superresolution
for ToF 3D Shape Scanning |
Sebastian
Schuon*, Stanford University
Christian Theobalt, Stanford University
James Davis, UC Santa Cruz
Sebastian Thrun, Stanford University |
Detecting Carried Objects from
Video Sequences |
Dima
Damen*, University of Leeds
David Hogg, University of Leeds |
Active Tracking of Two Free-Moving
Targets with a Stereo Head |
Luis
Perdigoto*, University of Coimbra
Joao Barreto, ISR
Rui Caseiro, ISR
Helder Araujo, ISR |
Observe Locally, Infer Globally: a
Space-Time MRF for Detecting Abnormal Activities with Incremental
Updates |
Jaechul
Kim*, UT-Austin
Kristen Grauman, UT-Austin |
Robust Scene Flow using Binocular
Stereo Sequences in Near-Real-Time |
Tobi
Vaudrey*, University of Auckland
Thomas Brox,
Clemens Rabe,
Andreas Wedel,
Uwe Franke, Daimler
Daniel Cremers |
From Structure-from-Motion Point
Clouds to Fast Location Recognition |
Arnold
Irschara*, TU Graz
Christopher Zach,
Jan-Michael Frahm, University of North Carolina at Chapel Hill
Horst Bischof, TU Graz |
Human Action Recognition with
Interest Points and Camera Motion Compensation |
Krystian
Mikolajczyk*, University of Surrey |
Stereo Matching with Nonparametric
Smoothness Priors in Feature Space |
Brandon
Smith*, University of Wisconsin-Madiso
Li Zhang, University of Wisconsin-Madison
Hailin Jin, Adobe Systems Incorporated |
Proposal
of Inside-Out Camera for Measuring 3D Gaze
Position in Free Space |
Kazuaki
Nishio, chubu university
Makoto Kimura, CREST, Japan Science and Technology Arency
Tomoyuki Nagahashi*, Chubu University
Hironobu Fujiyoshi, chubu university
Yutaka Hirata, chubu university |
Real-Time O(1) Bilateral Filtering |
Qingxiong
Yang*, University of Illinois, Urbana
Kar-Han Tan,
Narendra Ahuja, University of Illinois, Urbana |
Vision Based Mobile Mapping |
Frank
Verbiest*, Katholieke Universiteit Leuven
Luc Van Gool,
Maarten Vergauwen,
Marc Olijslagers, |
Markerless Motion Capture of
Skinned Models |
Luca
Ballan*, ETH Zurich |
Temporal Dithering of Illumination
for Fast Shape Acquisition |
Shuntaro
Yamazaki*, AIST
Koppal Sanjeev, CMU
Srinivasa Narasimhan, CMU |
Coplanar shadowgrams for acquiring
visual hulls of intricate objects |
Shuntaro
Yamazaki*, AIST
Srinivasa Narasimhan, CMU
Simon Baker, MSR
Takeo Kanade, CMU |
Learning General Optical Flow
Subspaces for Egomotion Estimation and Detection of Motion Anomalies |
Richard
Roberts*, Georgia Inst. of Technology
Christian Potthast, Georgia Inst. of Technology
Frank Dellaert, Georgia Inst. of Technology |
Sweetheart Detection in a Retail
Checkout Environment |
Quanfu
Fan*, IBM |
“Smart Room” with Real-time
Multi-camera People Tracking |
Kyungnam
Kim*, HRL Laboratories
Swarup Medasani,
Yuri Owechko, |
A Robust Approach for Automatic
Registration of Aerial Images with Untextured Aerial LiDAR Data |
Lu
Wang*, University of Southern California
Ulrich Neumann, USC |
Who should submit? (Sorry, the deadline has now passed!)
Everyone is welcome to submit a video. The CVPR Video Review is separate
from the papers, workshops, courses, and demos of the conference.
Supplementary materials submitted in those other categories are NOT
automatically part of the Video Review, and must be submitted here
explicitly to be considered. For example, authors of accepted papers are
encouraged to submit their work, including further or late-breaking
results that didn’t fit in the paper. Live demos or working systems are
another example, where the Video Review can serve as a record for those
who didn’t get to see your innovations in person. Working systems are also
very welcome. At least one of the authors must be registered for the
conference.
Important Dates
Submission deadline: 20th April 2009, at 16:00 US Pacific time (see world clock)
Decision Notification: 27th April 2009
General Instructions
Submissions will be assessed using these criteria:
- Visual appeal
- Technical content
- Informative / self-contained content
- Length commensurate with contribution (5 min. maximum)
- No product advertising, per-se
Videos must be submitted IN THEIR FINAL VERSION, as they should appear to the public. Reviewing will be single-blind. Please submit digital video files, with or without audio, encoded with a standard compression algorithm (see detailed instructions below) and in the resolution of 720×480 square pixels and 29.97 fps. Audio narrations are encouraged, but videos should not depend on them too much, since the venue may be noisy or the volume may get turned down during presentations.
Program Committee
Piotr Dollar, Caltech,
Mukta Prasad, ETH Zurich
Spencer Reynolds, ILM
Marc Sturzel, EADS
Drew Steedly, Microsoft LiveLabs
Mike Terry, U. of Waterloo
Detailed Instructions
MPEG-4 is an example of a fairly standard algorithm implemented in Quicktime 6+ or avi-friendly codecs such as xVid (preferred), DivX, and x264. 5 Minutes should fit within the submission site’s 100 Mb quota. These files will appear as .flv’s on the conference DVD, and will also be encoded to a video-DVD to play on NTSC monitors in Miami. If, instead, you are comfortable encoding with MPEG-2 for DVDs using tools such as TMPGEnc, then please keep the video bitrate below 7000 kbit/sec. For uniformity, each video MUST have a CVPR 2009 Video Review “leader” for 4+ seconds: a special static introduction sequence featuring the title, authors, and optional names/logos of the authors’ affiliations. The leader is provided here as a power-point.ppt or as a CVPR2009_VideoReviewLeader_blank.png, and looks like this:
Audio should be encoded with xVid, AC3, MP3, WAV/PCM, or AAC. Here is guitar5.wav to help standardize everyone’s audio-levels.
Further information & updates will appear here:
http://www.cvpr2009.org/call-for-videos/
Please direct questions to: cvpr09.videoreview@gmail.com
| |