-
公开(公告)号:US20200342301A1
公开(公告)日:2020-10-29
申请号:US16961932
申请日:2019-07-12
Inventor: Xiangshui MIAO , Yi LI , Wenqian PAN
Abstract: Disclosed by the disclosure is a convolutional neural network on-chip learning system based on non-volatile memory, comprising: an input module, a convolutional neural network module, an output module and a weight update module. The on-chip learning of the convolutional neural network module implements a synaptic function by using a characteristic which the conductance of a memristor changes according to an applied pulse, and the convolutional kernel value or synaptic weight value is stored in a memristor unit; the input module converts an input signal into a voltage signal required by the convolutional neural network module; the convolutional neural network module converts the input voltage signal level by level, and transmits the result to the output module to obtain an output of the network; and the weight update module adjusts the conductance value of the memristor in the convolutional neural network module according to the result of the output module to update a network convolutional kernel value or synaptic weight value.
-
公开(公告)号:US20200285954A1
公开(公告)日:2020-09-10
申请号:US16464977
申请日:2018-06-07
Inventor: Yi LI , Wenqian PAN , Xiangshui MIAO
Abstract: The present disclosure discloses a memory-based CNN, comprising: an input module, a convolution layer circuit module, a pooling layer circuit module, an activation function module, a fully connected layer circuit module, a softmax function module and an output module, convolution kernel values or synapse weights are stored in the NOR FLASH units; the input module converts an input signal into a voltage signal required by the convolutional neural network; the convolutional layer circuit module convolves the voltage signal corresponding to the input signal with the convolution kernel values, and transmits the result to the activation function module; the activation function module activates the signal; the pooling layer circuit module performs a pooling operation on the activated signal; the fully connected layer circuit module multiplies the pooled signal with the synapse weights to achieve classification; the softmax function module normalizes the classification result into probability values as an output of the entire network. The disclosure satisfies the requirements of real-time data processing and has low hardware cost.
-