Smart-DNN+: A Memory-efficient Neural Networks Compression Framework for the Model Inference.
Donglei WuWeihao YangXiangyu ZouWen XiaShiyi LiZhenbo HuWeizhe ZhangBinxing FangPublished in: ACM Trans. Archit. Code Optim. (2023)
Keyphrases
- probabilistic model
- neural network
- memory efficient
- theoretical framework
- formal model
- inference process
- prior knowledge
- mathematical model
- unified framework
- objective function
- pattern recognition
- mathematical framework
- data compression
- conceptual framework
- neural network model
- modeling framework
- bayesian framework
- image compression
- bayesian networks
- high level