Training quantized one-stage object detection neural networks via selective feature imitation.
Jiyuan JiaLi ZhouJie ChenPublished in: J. Electronic Imaging (2019)
Keyphrases
- object detection
- neural network
- training process
- training stage
- training algorithm
- feed forward neural networks
- feedforward neural networks
- back propagation
- multi layer perceptron
- backpropagation algorithm
- face detection
- neural network training
- error back propagation
- feature vectors
- pedestrian detection
- object recognition
- learning stage
- preprocessing stage
- scene understanding
- object categories
- object detectors
- feed forward
- supervised learning
- fuzzy logic
- image features
- training set
- reinforcement learning
- discriminatively trained
- multilayer perceptron
- test set
- artificial neural networks
- pattern recognition
- activation function
- neural nets
- neural network model
- radial basis function network
- fault diagnosis
- training examples
- feature extraction
- genetic algorithm