Differentiable Soft Quantization: Bridging Full-Precision and Low-Bit Neural Networks.
Ruihao GongXianglong LiuShenghu JiangTianxiang LiPeng HuJiazhen LinFengwei YuJunjie YanPublished in: ICCV (2019)
Keyphrases
- neural network
- uniform quantization
- pattern recognition
- objective function
- fuzzy logic
- fuzzy systems
- recurrent neural networks
- genetic algorithm
- artificial neural networks
- adaptive quantization
- network architecture
- high precision
- neural nets
- successive approximation
- activation function
- training process
- precision and recall
- multilayer perceptron
- neural network model
- loss function
- average precision
- back propagation
- quantization error
- motion estimation
- computational complexity
- support vector