Hardware-Aware Softmax Approximation for Deep Neural Networks.
Xue GengJie LinBin ZhaoAnmin KongMohamed M. Sabry AlyVijay ChandrasekharPublished in: ACCV (4) (2018)
Keyphrases
- neural network
- activation function
- hardware and software
- low cost
- real time
- multi layer
- pattern recognition
- computing power
- feed forward
- hardware implementation
- approximation algorithms
- error bounds
- artificial neural networks
- self organizing maps
- vlsi implementation
- high end
- approximation error
- neural nets
- recurrent neural networks
- computer systems
- fuzzy logic
- personal computer
- expert systems
- fuzzy systems
- multilayer perceptron
- genetic algorithm
- hidden layer
- fault diagnosis
- input output
- hardware design
- circuit design
- closed form
- machine learning
- massively parallel
- bayesian networks
- neural network model
- graphical models