TRQ: Ternary Neural Networks With Residual Quantization

Authors

  • Yue Li Beihang University
  • Wenrui Ding Beihang University
  • Chunlei Liu Beihang University
  • Baochang Zhang Beihang University
  • Guodong Guo Institute of Deep Learning,Baidu Research and National Engineering Laboratory for Deep Learning Technology and Application

Keywords:

(Deep) Neural Network Algorithms, Classification and Regression, Applications

Abstract

Ternary neural networks (TNNs) are potential for network acceleration by reducing the full-precision weights in network to ternary ones, e.g., {-1,0,1}. However, existing TNNs are mostly calculated based on rule-of-thumb quantization methods by simply thresholding operations, which causes a significant accuracy loss. In this paper, we introduce a stem-residual framework which provides new insight into Ternary quantization, termed Residual Quantization (TRQ), to achieve more powerful TNNs. Rather than directly thresholding operations, TRQ recursively performs quantization on full-precision weights for a refined reconstruction by combining the binarized stem and residual parts. With such a unique quantization process, TRQ endows the quantizer with high flexibility and precision. Our TRQ is generic, which can be easily extended to multiple bits through recursively encoded residual for a better recognition accuracy. Extensive experimental results demonstrate that the proposed method yields great recognition accuracy while being accelerated.

Downloads

Published

2021-05-18

How to Cite

Li, Y., Ding, W., Liu, C., Zhang, B., & Guo, G. (2021). TRQ: Ternary Neural Networks With Residual Quantization. Proceedings of the AAAI Conference on Artificial Intelligence, 35(10), 8538-8546. Retrieved from https://ojs.aaai.org/index.php/AAAI/article/view/17036

Issue

Section

AAAI Technical Track on Machine Learning III