Orthogonal Weight Normalization: Solution to Optimization Over Multiple Dependent Stiefel Manifolds in Deep Neural Networks

Authors

  • Lei Huang Beihang University
  • Xianglong Liu Beihang University
  • Bo Lang Beihang University
  • Adams Yu Carnegie Mellon University
  • Yongliang Wang JD.COM
  • Bo Li University of California, Berkeley

DOI:

https://doi.org/10.1609/aaai.v32i1.11768

Keywords:

Deep Learning, Neural Networks, Orthogonal matrix, Optimization, Stiefel Manifolds

Abstract

Orthogonal matrix has shown advantages in training Recurrent Neural Networks (RNNs), but such matrix is limited to be square for the hidden-to-hidden transformation in RNNs. In this paper, we generalize such square orthogonal matrix to orthogonal rectangular matrix and formulating this problem in feed-forward Neural Networks (FNNs) as Optimization over Multiple Dependent Stiefel Manifolds (OMDSM). We show that the orthogonal rectangular matrix can stabilize the distribution of network activations and regularize FNNs. We propose a novel orthogonal weight normalization method to solve OMDSM. Particularly, it constructs orthogonal transformation over proxy parameters to ensure the weight matrix is orthogonal. To guarantee stability, we minimize the distortions between proxy parameters and canonical weights over all tractable orthogonal transformations. In addition, we design orthogonal linear module (OLM) to learn orthogonal filter banks in practice, which can be used as an alternative to standard linear module. Extensive experiments demonstrate that by simply substituting OLM for standard linear module without revising any experimental protocols, our method improves the performance of the state-of-the-art networks, including Inception and residual networks on CIFAR and ImageNet datasets.

Downloads

Published

2018-04-29

How to Cite

Huang, L., Liu, X., Lang, B., Yu, A., Wang, Y., & Li, B. (2018). Orthogonal Weight Normalization: Solution to Optimization Over Multiple Dependent Stiefel Manifolds in Deep Neural Networks. Proceedings of the AAAI Conference on Artificial Intelligence, 32(1). https://doi.org/10.1609/aaai.v32i1.11768