Spiking neural network based on joint entropy of optical flow features for human action recognition.
S. Jeba BerlinMala JohnPublished in: Vis. Comput. (2022)
Keyphrases
- optical flow
- feature extraction
- computer vision
- low level
- least squares
- action recognition
- feed forward
- neural network
- image sequences
- feature space
- feature vectors
- information theory
- motion estimation
- information theoretic
- single neuron
- extracted features
- information content
- image features
- moving objects
- feature selection