Deep Convolutional Neural Network Inference with Floating-point Weights and Fixed-point Activations.
Liangzhen LaiNaveen SudaVikas ChandraPublished in: CoRR (2017)
Keyphrases
- floating point
- fixed point
- convolutional neural network
- face detection
- sufficient conditions
- sparse matrices
- neural network
- bayesian networks
- fixed point theorem
- instruction set
- floating point arithmetic
- constraint databases
- belief propagation
- dynamical systems
- free energy
- policy iteration
- post processing
- computer vision