Multi-level Residual Networks from Dynamical Systems ViewDownload PDF

15 Feb 2018 (modified: 10 Feb 2022)ICLR 2018 Conference Blind SubmissionReaders: Everyone
Abstract: Deep residual networks (ResNets) and their variants are widely used in many computer vision applications and natural language processing tasks. However, the theoretical principles for designing and training ResNets are still not fully understood. Recently, several points of view have emerged to try to interpret ResNet theoretically, such as unraveled view, unrolled iterative estimation and dynamical systems view. In this paper, we adopt the dynamical systems point of view, and analyze the lesioning properties of ResNet both theoretically and experimentally. Based on these analyses, we additionally propose a novel method for accelerating ResNet training. We apply the proposed method to train ResNets and Wide ResNets for three image classification benchmarks, reducing training time by more than 40\% with superior or on-par accuracy.
Keywords: residual networks, dynamical systems
Data: [CIFAR-10](https://paperswithcode.com/dataset/cifar-10), [CIFAR-100](https://paperswithcode.com/dataset/cifar-100), [STL-10](https://paperswithcode.com/dataset/stl-10)
9 Replies

Loading