MUSCLE-VCD-2007
| Domain | Video archives |
| Media | Video |
| Size | 10 Gb |
| Instances | 100 hours |
| File Format | MPG |
| Creation Date | 2007 |
| Task | content-based copy detection |
| Copyright | Research purpose only |
| URL | http://www-rocq.inria.fr/imedia/civr-bench/data.html |
Description
MUSCLE-VCD-2007 is dedicated to the evaluation of content-based video detection systems. It contains about 100 hours of video materials coming from different sources: web video clips, TV archives, movies and some video queries to be retrieved (2 hours 30 mins). The videos cover very large kind of programs: documentaries, movies, sports events, TV shows, cartoons etc.
Quality
The videos have different bitrates, different resolutions and different video format. These videos have been provided in their original format and also in an MPEG1 format by a reencoding.
Source
The corpus was used for CIVR 2007 video copy detection evaluation showcase. It was crated in the context of MUSCLE EU network of excellence.
Ground Truth Annotation
Ground truth annotations are provided for 15 queries (transformed videos).
Features
No additional features
Licensing / Copyright
The corpus is for research purposes only. The origin of the corpus must be precised in any publication related to its use: MUSCLE-VCD-2007, http://www-rocq.inria.fr/imedia/civr-bench/index.html. The videos themselves come from two websites www.archive.org and www.open-video.org The terms of use can be found here: archive.org terms of use, open-video.org terms of use.
Citation
MUSCLE-VCD-2007
