Interpretable Basis Decomposition for Visual Explanation
Author(s)
Zhou, Bolei; Sun, Yiyou; Torralba, Antonio; Bau, David
DownloadAccepted version (5.108Mb)
Terms of use
Metadata
Show full item recordAbstract
Explanations of the decisions made by a deep neural network are important for human end-users to be able to understand and diagnose the trustworthiness of the system. Current neural networks used for visual recognition are generally used as black boxes that do not provide any human interpretable justification for a prediction. In this work we propose a new framework called Interpretable Basis Decomposition for providing visual explanations for classification networks. By decomposing the neural activations of the input image into semantically interpretable components pre-trained from a large concept corpus, the proposed framework is able to disentangle the evidence encoded in the activation feature vector, and quantify the contribution of each piece of evidence to the final prediction. We apply our framework for providing explanations to several popular networks for visual recognition, and show it is able to explain the predictions given by the networks in a human-interpretable way. The human interpretability of the visual explanations provided by our framework and other recent explanation methods is evaluated through Amazon Mechanical Turk, showing that our framework generates more faithful and interpretable explanations (The code and data are available at https://github.com/CSAILVision/IBD).
Date issued
2018-10Department
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory; Massachusetts Institute of Technology. Laboratory for Computer ScienceJournal
European Conference on Computer Vision
Publisher
Springer International Publishing
Citation
Zhou, Bolei et al. "Interpretable Basis Decomposition for Visual Explanation." European Conference on Computer Vision, September 2018, Munich, Germany, Springer Nature, October 2018 © 2018 Springer Nature
Version: Author's final manuscript
ISBN
9783030012366
9783030012373
ISSN
0302-9743
1611-3349