Pooling the Convolutional Layers in Deep ConvNets for Action Recognition.
Shichao ZhaoYanbin LiuYahong HanRichang HongPublished in: CoRR (2015)
Keyphrases
- action recognition
- deep belief networks
- feed forward
- deep learning
- restricted boltzmann machine
- bag of words
- human detection
- human actions
- probabilistic model
- activity recognition
- generative model
- computer vision
- machine learning
- sparse coding
- spatial temporal
- back propagation
- action classification
- static images
- independent subspace analysis
- artificial neural networks
- body parts
- recognition of human actions
- neural network
- mid level
- recognizing actions
- video dataset
- depth cameras
- bag of features
- recognizing human actions
- human activities
- learning algorithm
- action primitives
- max margin
- action recognition in videos
- view invariant action recognition