SDAN: Stacked Diverse Attention Network for Video Action Recognition.
Xiaoguang ZhuSiran HuangWenjing FanYuhao ChengHuaqing ShaoPeilin LiuPublished in: ISCAS (2021)
Keyphrases
- action recognition
- human actions
- action classification
- spatial temporal
- video dataset
- action detection
- recognizing human actions
- bag of words
- static images
- motion features
- computer vision
- human activities
- space time interest points
- recognition of human actions
- activity recognition
- spatio temporal interest points
- human detection
- body parts
- view invariant
- mid level
- motion history images
- multimedia
- video data
- human pose
- motion capture data
- three dimensional
- video sequences
- bag of features
- space time
- recognizing actions
- object recognition
- human motion
- event detection
- video content
- action recognition in videos
- temporal information
- video surveillance
- human activity recognition
- video streams
- video images
- video frames