Forecasting Time-to-Collision from Monocular Video: Feasibility, Dataset, and Challenges.
Aashi ManglikXinshuo WengEshed Ohn-BarKris M. KitaniPublished in: IROS (2019)
Keyphrases
- human actions
- image sequences
- video sequences
- short term
- multimedia
- lessons learned
- collision detection
- pose estimation
- video content
- video data
- video frames
- video streams
- video database
- video clips
- weakly labeled
- real time
- video surveillance
- real world
- action recognition
- multimedia data
- activity recognition
- video analysis
- video dataset
- spatio temporal
- articulated human motion
- video retrieval
- space time
- event recognition
- collision avoidance
- benchmark datasets
- video annotation
- trecvid multimedia event detection