CCR: A concise convolution rule for sparse neural network accelerators.
Jiajun LiGuihai YanWenyan LuShuhao JiangShijun GongJingya WuXiaowei LiPublished in: DATE (2018)
Keyphrases
- neural network
- neural network model
- back propagation
- nearest neighbour
- pattern recognition
- neural network is trained
- artificial neural networks
- sparse data
- image processing
- sparse representation
- feed forward
- feed forward neural networks
- single chip
- rule sets
- compressive sensing
- computing systems
- activation function
- recurrent neural networks
- neural nets
- hidden layer
- network architecture
- fuzzy logic
- fermentation process
- learning vector quantization
- concise representation
- convolution kernel
- low power
- sparse coding
- classification rules
- linear combination
- signal processing
- association rules
- genetic algorithm