A Theoretical Explanation for Perplexing Behaviors of Backpropagation-based Visualizations
Abstract
Backpropagation-based visualizations have been proposed to interpret convolutional neural networks (CNNs), however a theory is missing to justify their behaviors: Guided backpropagation (GBP) and deconvolutional network (DeconvNet) generate more human-interpretable but less class-sensitive visualizations than saliency map. Motivated by this, we develop a theoretical explanation revealing that GBP and DeconvNet are essentially doing (partial) image recovery which is unrelated to the network decisions. Specifically, our analysis shows that the backward ReLU introduced by GBP and DeconvNet, and the local connections in CNNs are the two main causes of compelling visualizations. Extensive experiments are provided that support the theoretical analysis.
Cite
@article{arxiv.1805.07039,
title = {A Theoretical Explanation for Perplexing Behaviors of Backpropagation-based Visualizations},
author = {Weili Nie and Yang Zhang and Ankit Patel},
journal= {arXiv preprint arXiv:1805.07039},
year = {2020}
}
Comments
21 pages, ICML 2018 (We revised the proofs of Theorem 1 and 2 in Appendix)