UNPU: An Energy-Efficient Deep Neural Network Accelerator With Fully Variable Weight Bit Precision.
Jinmook LeeChanghyeon KimSanghoon KangDongjoo ShinSangyeob KimHoi-Jun YooPublished in: IEEE J. Solid State Circuits (2019)
Keyphrases
- neural network
- wireless sensor networks
- weight update
- neural network model
- energy efficient
- pattern recognition
- artificial neural networks
- sensor networks
- back propagation
- genetic algorithm
- neural nets
- prediction model
- precision and recall
- feed forward
- high precision
- network model
- parallel implementation
- multi layer perceptron
- deep learning
- self organizing maps
- multi layer
- energy consumption
- neural network is trained
- image reconstruction from projections