TVE: Learning Meta-attribution for Transferable Vision Explainer
Document Type
Conference Proceeding
Publication Date
1-1-2024
Abstract
Explainable machine learning significantly improves the transparency of deep neural networks. However, existing work is constrained to explaining the behavior of individual model predictions, and lacks the ability to transfer the explanation across various models and tasks. This limitation results in explaining various tasks being time- and resource-consuming. To address this problem, we introduce a Transferable Vision Explainer (TVE) that can effectively explain various vision models in downstream tasks. Specifically, the transferability of TVE is realized through a pre-training process on large-scale datasets towards learning the meta-attribution. This meta-attribution leverages the versatility of generic backbone encoders to comprehensively encode the attribution knowledge for the input instance, which enables TVE to seamlessly transfer to explain various downstream tasks, without the need for training on task-specific data. Empirical studies involve explaining three different architectures of vision models across three diverse downstream datasets. The experimental results indicate TVE is effective in explaining these tasks without the need for additional training on downstream data. The source code is available at https://github.com/guanchuwang/TVE.
Identifier
85203845404 (Scopus)
Publication Title
Proceedings of Machine Learning Research
e-ISSN
26403498
First Page
50248
Last Page
50267
Volume
235
Grant
IIS-1900990
Fund Ref
National Institutes of Health
Recommended Citation
Wang, Guanchu; Chuang, Yu Neng; Yang, Fan; Du, Mengnan; Chang, Chia Yuan; Zhong, Shaochen; Liu, Zirui; Xu, Zhaozhuo; Zhou, Kaixiong; Cai, Xuanting; and Hu, Xia, "TVE: Learning Meta-attribution for Transferable Vision Explainer" (2024). Faculty Publications. 878.
https://digitalcommons.njit.edu/fac_pubs/878
