Dataflow Restructuring for Active Memory Reduction in Deep Neural Networks.
Antonio CipollettaAndrea CalimeraPublished in: DATE (2021)
Keyphrases
- neural network
- auto associative
- pattern recognition
- associative memory
- genetic algorithm
- artificial neural networks
- fuzzy logic
- memory usage
- back propagation
- low memory
- memory size
- control flow
- data flow
- multilayer perceptron
- parallel computing
- training process
- activation function
- computing power
- limited memory
- neural network model
- main memory
- design methodology
- hidden layer
- neural nets
- feed forward
- memory requirements
- deep learning
- cellular neural networks
- recently developed
- fault diagnosis
- neural network is trained