Wide Residual Networks

TLDR

Deep residual networks can scale to thousands of layers with improving performance, but each percent of accuracy gain nearly doubles layers, leading to diminishing feature reuse and slow training. The authors aim to address training inefficiencies in deep residual networks by experimentally studying ResNet block designs and proposing a new architecture that reduces depth while increasing width. They experimentally analyze ResNet block designs and introduce wide residual networks that are shallower but wider. Wide residual networks outperform thin, very deep counterparts, with a 16‑layer WRN achieving state‑of‑the‑art accuracy and efficiency on CIFAR, SVHN, COCO, and ImageNet, and the authors provide code and models online.

Abstract

Deep residual networks were shown to be able to scale up to thousands of layers and still have improving performance. However, each fraction of a percent of improved accuracy costs nearly doubling the number of layers, and so training very deep residual networks has a problem of diminishing feature reuse, which makes these networks very slow to train. To tackle these problems, in this paper we conduct a detailed experimental study on the architecture of ResNet blocks, based on which we propose a novel architecture where we decrease depth and increase width of residual networks. We call the resulting network structures wide residual networks (WRNs) and show that these are far superior over their commonly used thin and very deep counterparts. For example, we demonstrate that even a simple 16-layer-deep wide residual network outperforms in accuracy and efficiency all previous deep residual networks, including thousand-layer-deep networks, achieving new state-of-the-art results on CIFAR, SVHN, COCO, and significant improvements on ImageNet. Our code and models are available at this https URL