Learnable Visual Rhythms Based on the Stacking of Convolutional Neural Networks for Action Recognition.
Helena Almeida MaiaMarcos Roberto e SouzaAnderson SantosHélio PedriniHemerson TaconAndré de Souza BritoHugo de Lima ChavesMarcelo Bernardes VieiraSaulo Moraes VillelaPublished in: ICMLA (2019)
Keyphrases
- action recognition
- convolutional neural networks
- mid level
- human actions
- bag of words
- activity recognition
- action classification
- visual information
- human detection
- spatial temporal
- convolutional network
- body parts
- computer vision
- recognizing human actions
- view invariant
- low level
- bag of features
- human activities
- visual features
- recognition of human actions
- static images
- motion features
- video dataset
- learning algorithm
- high level
- recognizing actions
- action primitives
- independent subspace analysis
- depth sensors
- depth cameras
- human activity recognition
- human pose
- multiscale