Pooling the Convolutional Layers in Deep ConvNets for Video Action Recognition.
Shichao ZhaoYanbin LiuYahong HanRichang HongQinghua HuQi TianPublished in: IEEE Trans. Circuits Syst. Video Technol. (2018)
Keyphrases
- action recognition
- human actions
- action classification
- feed forward
- video dataset
- spatial temporal
- action detection
- deep belief networks
- recognizing human actions
- static images
- motion features
- recognition of human actions
- human activities
- deep learning
- bag of words
- activity recognition
- human detection
- computer vision
- space time interest points
- multimedia
- video sequences
- mid level
- video streams
- back propagation
- body parts
- motion history images
- depth sensors
- video data
- human pose
- bag of features
- probabilistic model
- view invariant
- recognizing actions
- neural network
- bag of visual words
- space time
- restricted boltzmann machine
- video surveillance
- video images
- image representation
- spatio temporal
- key frames
- video clips
- action recognition in videos