Towards Black-Box Explainability with Gaussian Discriminant Knowledge Distillation
Abstract
In this paper, we propose a method for post-hoc ex-plainability of black-box models. The key component of the semantic and quantitative local explanation is a knowledge distillation (KD) process which is used to mimic the teacher’s behavior by means of an explainable generative model. Therefore, we introduce a Concept Probability Density Encoder (CPDE) in conjunction with a Gaussian Discriminant Decoder (GDD) to describe the contribution of high-level concepts (e.g. object parts, color, shape). We argue that our objective function encourages both, an explanation given by a set of likelihood ratios and a measure to describe how far the explainer deviates from the training data distribution of the concepts. The method can leverage any pre-trained concept classifier that admits concept scores (e.g. logits) or probabilities. We demonstrate the effectiveness of the proposed method in the context of object detection utilizing the DensePose dataset.
Cite
Text
Haselhoff et al. "Towards Black-Box Explainability with Gaussian Discriminant Knowledge Distillation." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2021. doi:10.1109/CVPRW53098.2021.00011Markdown
[Haselhoff et al. "Towards Black-Box Explainability with Gaussian Discriminant Knowledge Distillation." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2021.](https://mlanthology.org/cvprw/2021/haselhoff2021cvprw-blackbox/) doi:10.1109/CVPRW53098.2021.00011BibTeX
@inproceedings{haselhoff2021cvprw-blackbox,
title = {{Towards Black-Box Explainability with Gaussian Discriminant Knowledge Distillation}},
author = {Haselhoff, Anselm and Kronenberger, Jan and Küppers, Fabian and Schneider, Jonas},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
year = {2021},
pages = {21-28},
doi = {10.1109/CVPRW53098.2021.00011},
url = {https://mlanthology.org/cvprw/2021/haselhoff2021cvprw-blackbox/}
}