Neural network acceleration methods via selective activation

The increase in neural network recognition accuracy is accompanied by a significant increase in the scales of networks and computations. To make deep learning frameworks widely used on mobile platforms, model acceleration has become extremely important in computer vision. In this study, a novel neural network acceleration method based on selective activation is proposed. First, as the algebraic basis for selective activation, mask general matrix multiplication is used to reduce matrix multiplication calculations. Second, to screen and remove activated neurons and reduce the number of calculations, we introduce an Activation Management Unit that includes two different strategies, Selective Activation with Primary Weights (SAPW) and Selective Activation with Primary Inputs (SAPI). SAPW greatly reduces the number of calculations of the fully connected layer and self-attention and better guarantees detection accuracy. SAPI has the best performance on convolutional architectures, which can significantly reduce the amount of convolutional computation while maintaining the image classification accuracy. We present result of extensive experiments on computational and accuracy tradeoffs and show strong performance for CIFAR-10 classification and Pascal VOC2012 detection. Compared with the dense method, the proposed selective activation method significantly reduces the number of neural network calculations with equal accuracy.

Automated summary

In this study, a novel neural network acceleration method based on selective activation is proposed. By using selective activation as the algebraic basis to reduce matrix multiplication calculations, and introducing an Activation Management Unit to screen and remove activated neurons and reduce the number of calculations, this method significantly reduces the number of neural network calculations while maintaining accuracy.