MSP: An FPGA-Specific Mixed-Scheme, Multi-Precision Deep Neural Network Quantization Framework.
Sung-En ChangYanyu LiMengshu SunWeiwen JiangRunbin ShiXue LinYanzhi WangPublished in: CoRR (2020)
Keyphrases
- neural network
- main contribution
- theoretical framework
- representation scheme
- high precision
- genetic algorithm
- real time
- learning scheme
- artificial neural networks
- fpga implementation
- hardware implementation
- neural network model
- domain specific
- high speed
- probabilistic model
- control system
- video sequences
- multiscale
- learning algorithm