Fu, W, Breininger, K, Schaffert, R et al. (2 more authors) (2019) A Divide-and-Conquer Approach Towards Understanding Deep Networks. In: Lecture Notes in Computer Science. MICCAI 2019: International Conference on Medical Image Computing and Computer-Assisted Intervention, 13-17 Oct 2019, Shenzen, China. Springer Verlag , pp. 183-191. ISBN 9783030322380
Abstract
Deep neural networks have achieved tremendous success in various fields including medical image segmentation. However, they have long been criticized for being a black-box, in that interpretation, understanding and correcting architectures is difficult as there is no general theory for deep neural network design. Previously, precision learning was proposed to fuse deep architectures and traditional approaches. Deep networks constructed in this way benefit from the original known operator, have fewer parameters, and improved interpretability. However, they do not yield state-of-the-art performance in all applications. In this paper, we propose to analyze deep networks using known operators, by adopting a divide-and-conquer strategy to replace network components, whilst retaining networks performance. The task of retinal vessel segmentation is investigated for this purpose. We start with a high-performance U-Net and show by step-by-step conversion that we are able to divide the network into modules of known operators. The results indicate that a combination of a trainable guided filter and a trainable version of the Frangi filter yields a performance at the level of U-Net (AUC 0.974 vs. 0.972) with a tremendous reduction in parameters (111, 536 vs. 9, 575). In addition, the trained layers can be mapped back into their original algorithmic interpretation and analyzed using standard tools of signal processing.
Metadata
Item Type: | Proceedings Paper |
---|---|
Authors/Creators: |
|
Copyright, Publisher and Additional Information: | © Springer Nature Switzerland AG 2019. This is an author produced version of a conference paper published in Lecture Notes in Computer Science. Uploaded in accordance with the publisher's self-archiving policy. |
Keywords: | Precision learning; Debugging; CNN |
Dates: |
|
Institution: | The University of Leeds |
Academic Units: | The University of Leeds > Faculty of Engineering & Physical Sciences (Leeds) > School of Computing (Leeds) |
Depositing User: | Symplectic Publications |
Date Deposited: | 16 Apr 2020 13:38 |
Last Modified: | 13 Jul 2020 15:37 |
Status: | Published |
Publisher: | Springer Verlag |
Identification Number: | 10.1007/978-3-030-32239-7_21 |
Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:159512 |