SSVOD: Semi-Supervised Video Object Detection with Sparse Annotations.
Tanvir MahmudChun-Hao LiuBurhaneddin YamanDiana MarculescuPublished in: WACV (2024)
Keyphrases
- object detection
- semi supervised
- multimedia
- natural language descriptions
- video data
- video content
- video material
- semi supervised learning
- video surveillance
- action detection
- video streams
- video analysis
- real time face tracking
- high dimensional
- image annotation
- object categories
- metadata
- video sequences
- computer vision
- unsupervised learning
- video clips
- multi view
- real time
- object recognition
- key frames
- object class
- pedestrian detection
- user generated
- semantic annotation
- multimedia data
- background subtraction
- video frames
- space time
- face detection
- unlabeled data
- constrained clustering
- semi supervised classification
- partially labeled
- pairwise constraints
- human activities
- human detection
- scene understanding
- video retrieval
- sparse coding
- sparse representation
- visual features
- active learning
- keywords
- learning algorithm