# Explaining image classifiers by removing input features using generative models

arXiv: https://arxiv.org/abs/1910.04256 · HTML: https://mlsanity.com/paper/1910.04256/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 9 Oct 2019.

## Built on by (most cited first, 5 of 5)

- [Abs-CAM: A Gradient Optimization Interpretable Approach for Explanation of Convolutional Neural Networks](https://mlsanity.com/paper/2207.03648/) (2022)
- [Generative Counterfactuals for Neural Networks via Attribute-Informed Perturbation](https://mlsanity.com/paper/2101.06930/) (2021)
- [Deconfounding to Explanation Evaluation in Graph Neural Networks](https://mlsanity.com/paper/2201.08802/) (2022)
- [Explaining the Black-box Smoothly- A Counterfactual Approach](https://mlsanity.com/paper/2101.04230/) (2021)
- [Understanding CNNs from excitations](https://mlsanity.com/paper/2205.00932/) (2022)

## Abstract (arXiv)

Perturbation-based explanation methods often measure the contribution of an input feature to an image classifier's outputs by heuristically removing it via e.g. blurring, adding noise, or graying out, which often produce unrealistic, out-of-samples. Instead, we propose to integrate a generative inpainter into three representative attribution methods to remove an input feature. Our proposed change improved all three methods in (1) generating more plausible counterfactual samples under the true data distribution; (2) being more accurate according to three metrics: object localization, deletion, and saliency metrics; and (3) being more robust to hyperparameter changes. Our findings were consistent across both ImageNet and Places365 datasets and two different pairs of classifiers and inpainters.