DiscrimNet: Semi-Supervised Action Recognition from Videos using Generative Adversarial Networks.
Unaiza AhsanChen SunIrfan A. EssaPublished in: CoRR (2018)
Keyphrases
- action recognition
- human actions
- semi supervised
- action classification
- video dataset
- recognition of human actions
- recognizing human actions
- ucf sports
- spatio temporal interest points
- view invariant
- static images
- recognizing actions
- activity recognition
- human activities
- generative model
- bag of words
- action detection
- human detection
- motion features
- space time interest points
- action recognition in videos
- mid level features
- computer vision
- motion history images
- body parts
- semi supervised learning
- human object interactions
- spatial temporal
- motion capture data
- action primitives
- labeled data
- active learning
- human pose
- depth sensors
- multi view
- spatio temporal
- video sequences
- bag of features
- video data
- machine learning