VideoLSTM convolves, attends and flows for action recognition.
Zhenyang LiKirill GavrilyukEfstratios GavvesMihir JainCees G. M. SnoekPublished in: Comput. Vis. Image Underst. (2018)
Keyphrases
- action recognition
- bag of words
- activity recognition
- human actions
- human detection
- body parts
- computer vision
- action classification
- bag of features
- static images
- independent subspace analysis
- spatial temporal
- human activities
- mid level
- recognizing human actions
- recognizing actions
- view invariant
- action recognition in videos
- action primitives
- view invariant action recognition
- depth sensors
- video dataset
- object recognition