English

Random Bias Initialization Improves Quantized Training

Machine Learning 2020-04-22 v2 Computer Vision and Pattern Recognition Neural and Evolutionary Computing

Abstract

Binary neural networks improve computationally efficiency of deep models with a large margin. However, there is still a performance gap between a successful full-precision training and binary training. We bring some insights about why this accuracy drop exists and call for a better understanding of binary network geometry. We start with analyzing full-precision neural networks with ReLU activation and compare it with its binarized version. This comparison suggests to initialize networks with random bias, a counter-intuitive remedy.

Keywords

Cite

@article{arxiv.1909.13446,
  title  = {Random Bias Initialization Improves Quantized Training},
  author = {Xinlin Li and Vahid Partovi Nia},
  journal= {arXiv preprint arXiv:1909.13446},
  year   = {2020}
}