Rekall: Specifying Video Events using Compositions of Spatiotemporal Labels.
Daniel Y. FuWill CrichtonJames HongXinwei YaoHaotian ZhangAnh TruongAvanika NarayanManeesh AgrawalaChristopher RéKayvon FatahalianPublished in: CoRR (2019)
Keyphrases
- space time
- event recognition
- event detection
- spatial and temporal
- video clips
- video event
- spatiotemporal features
- surveillance videos
- video data
- video content
- human activities
- video sequences
- temporal information
- spatiotemporal patterns
- multimedia
- video analysis
- video surveillance
- weakly labeled
- real time
- temporal structure
- video database
- dynamic textures
- video streams
- video frames
- tv news
- event models
- video retrieval
- moving objects
- training data
- spatio temporal
- real time video
- temporal relationships
- soccer video
- pairwise
- temporal sequences
- video dataset
- temporal patterns
- temporal events
- event sequences
- video event detection
- dynamic scenes
- multi label
- temporal relations
- semantic labels
- training set
- video segments
- digital video