REACT: a heterogeneous reconfigurable neural network accelerator with software-configurable NoCs for training and inference on wearables.
Mohit UpadhyayRohan JunejaBo WangJun ZhouWeng-Fai WongLi-Shiuan PehPublished in: DAC (2022)
Keyphrases
- neural network
- training algorithm
- training process
- computer software
- feed forward neural networks
- structured prediction
- feedforward neural networks
- multi layer perceptron
- neural network training
- training patterns
- software systems
- training phase
- train a neural network
- software development
- back propagation
- artificial neural networks
- instructor led
- genetic algorithm
- human resource development
- field programmable gate array
- low cost
- training data
- training set
- parallel implementation
- self organizing maps
- test cases
- support vector
- neural network structure
- fuzzy logic
- backpropagation algorithm
- recurrent networks
- general purpose
- computer systems
- training examples
- bp neural network
- fault diagnosis
- radial basis function
- neural network model
- hidden layer
- feed forward
- multi layer
- probabilistic inference
- neural nets
- embedded systems