A semi-automatic tool for detection and tracking ground truth generation in videos.
Isaak KavasidisSimone PalazzoRoberto Di SalvoDaniela GiordanoConcetto SpampinatoPublished in: VIGTA@AVI (2012)
Keyphrases
- semi automatic
- ground truth
- gold standard
- landmark extraction
- hand detection
- fully automatic
- design rationale
- video scene
- human detection
- semi automatically
- video surveillance
- ontology development
- video sequences
- ontology mapping
- domain ontology
- labor intensive
- automatic detection
- crowded scenes
- detecting and tracking multiple
- particle filter
- partial occlusion
- semantic annotation
- object tracking
- wrapper generation
- surveillance videos
- detection algorithm
- high quality
- event detection
- video dataset
- domain knowledge
- knowledge acquisition
- ontology construction
- video data
- video analysis
- mean shift
- fully automated