Multi-Scale Dynamic Fixed-Point Quantization and Training for Deep Neural Networks.
Po-Yuan ChenHung-Che LinJiun-In GuoPublished in: ISCAS (2023)
Keyphrases
- fixed point
- neural network
- multiscale
- cellular neural networks
- training process
- sufficient conditions
- lookup table
- floating point
- constraint databases
- bargaining solution
- back propagation
- dynamical systems
- variational inequalities
- fixed point theorem
- multi layer perceptron
- belief propagation
- scale space
- multilayer perceptron
- semi supervised
- artificial neural networks
- image retrieval
- object recognition
- similarity measure
- three dimensional
- image processing