Risky Action Recognition in Lane Change Video Clips using Deep Spatiotemporal Networks with Segmentation Mask Transfer.
Ekim YurtseverYongkang LiuJacob LambertChiyomi MiyajimaEijiro TakeuchiKazuya TakedaJohn H. L. HansenPublished in: ITSC (2019)
Keyphrases
- action recognition
- video clips
- video database
- video dataset
- human actions
- bag of words
- computer vision
- activity recognition
- video streams
- video shots
- video data
- video retrieval
- video frames
- human pose
- action classification
- body parts
- video scene
- spatial and temporal
- recognizing human actions
- recognition of human actions
- temporal information
- mid level
- key frames
- object segmentation
- recognizing actions
- view invariant
- video content
- video segments
- moving objects
- video sequences
- action recognition in videos
- background subtraction
- space time
- image sequences
- three dimensional