Towards Faster Training of Global Covariance Pooling Networks by Iterative Matrix Square Root Normalization

TL;DR

This paper introduces an iterative matrix square root normalization method for global covariance pooling in CNNs, significantly speeding up training and achieving state-of-the-art results on large-scale and fine-grained image benchmarks.

Contribution

It proposes a novel iterative normalization technique using a meta-layer with nonlinear structured layers, enabling faster GPU-compatible training of covariance pooling networks.

Findings

Faster training compared to eigendecomposition-based methods.

Achieves competitive performance on ImageNet.

Establishes state-of-the-art results on fine-grained benchmarks.

Abstract

Global covariance pooling in convolutional neural networks has achieved impressive improvement over the classical first-order pooling. Recent works have shown matrix square root normalization plays a central role in achieving state-of-the-art performance. However, existing methods depend heavily on eigendecomposition (EIG) or singular value decomposition (SVD), suffering from inefficient training due to limited support of EIG and SVD on GPU. Towards addressing this problem, we propose an iterative matrix square root normalization method for fast end-to-end training of global covariance pooling networks. At the core of our method is a meta-layer designed with loop-embedded directed graph structure. The meta-layer consists of three consecutive nonlinear structured layers, which perform pre-normalization, coupled matrix iteration and post-compensation, respectively. Our method is much faster than EIG or SVD based ones, since it involves only matrix multiplications, suitable for parallel implementation on GPU. Moreover, the proposed network with ResNet architecture can converge in much less epochs, further accelerating network training. On large-scale ImageNet, we achieve competitive performance superior to existing counterparts. By finetuning our models pre-trained on ImageNet, we establish state-of-the-art results on three challenging fine-grained benchmarks. The source code and network models will be available at http://www.peihuali.org/iSQRT-COV