XS-VID: An Extremely Small Video Object Detection Dataset.
Jiahao GuoZiyang XuLianjun WuFei GaoWenyu LiuXinggang WangPublished in: CoRR (2024)
Keyphrases
- object detection
- pascal voc
- multimedia
- object detectors
- human actions
- video sequences
- video streams
- real time
- weakly labeled
- video data
- video content
- video frames
- computer vision
- small number
- face detection
- object recognition
- video images
- trecvid multimedia event detection
- video analysis
- pedestrian detection
- object categories
- action recognition
- video clips
- video retrieval
- object class
- video database
- object classification
- background subtraction
- video processing
- scene recognition
- event recognition
- video dataset
- online video
- action detection
- generative model
- database