Two-Stream Architecture Using RGB-based ConvNet and Pose-based LSTM for Video Action Recognition.
Ching-Jung HuangMunkhjargal GochooTan-Hsu TanPublished in: IIT (2023)
Keyphrases
- action recognition
- action detection
- human actions
- action classification
- human pose
- recognizing actions
- video dataset
- spatial temporal
- real time
- static images
- bag of words
- recognizing human actions
- activity recognition
- recognition of human actions
- motion features
- human activities
- computer vision
- human detection
- body parts
- space time interest points
- data streams
- multimedia
- mid level
- video data
- pose estimation
- motion history images
- video streams
- color space
- color images
- video surveillance
- human motion
- action recognition in videos
- bag of features
- video sequences
- machine learning
- human computer interaction
- space time
- object tracking
- color information
- key frames
- video retrieval