Residual Networks of Residual Networks: Multilevel Residual Networks
This work addresses the problem of improving optimization in deep neural networks for computer vision researchers, offering a novel architecture that boosts performance across multiple datasets and network types.
The paper tackles the optimization limitations of deep residual networks by proposing Residual Networks of Residual Networks (RoR), which adds level-wise shortcut connections to enhance learning capability, achieving state-of-the-art results with test errors of 3.77% on CIFAR-10, 19.73% on CIFAR-100, and 1.59% on SVHN.
A residual-networks family with hundreds or even thousands of layers dominates major image recognition tasks, but building a network by simply stacking residual blocks inevitably limits its optimization ability. This paper proposes a novel residual-network architecture, Residual networks of Residual networks (RoR), to dig the optimization ability of residual networks. RoR substitutes optimizing residual mapping of residual mapping for optimizing original residual mapping. In particular, RoR adds level-wise shortcut connections upon original residual networks to promote the learning capability of residual networks. More importantly, RoR can be applied to various kinds of residual networks (ResNets, Pre-ResNets and WRN) and significantly boost their performance. Our experiments demonstrate the effectiveness and versatility of RoR, where it achieves the best performance in all residual-network-like structures. Our RoR-3-WRN58-4+SD models achieve new state-of-the-art results on CIFAR-10, CIFAR-100 and SVHN, with test errors 3.77%, 19.73% and 1.59%, respectively. RoR-3 models also achieve state-of-the-art results compared to ResNets on ImageNet data set.