Research map / Label-efficient and robust learning / Recognition and long-tail
EPIC: Explanation of Pretrained Image Classification Networks via Prototypes
Piotr Borycki, Magdalena Trędowicz, Szymon Janusz, Jacek Tabor, Przemysław Spurek, Arkadiusz Lewicki, Łukasz Struski
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| plug the EPIC disentanglement module into a pretrained classifier | pretrained ImageNet classifiers · ImageNet | accuracy: ResNet-50 80.8% -> 80.8% | No clear effect |
| use EPIC instead of InfoDisent (finetuned head) | ImageNet classifiers · ImageNet | accuracy: 67.8% -> 80.8% (ResNet-50) | Improved |
| show EPIC prototype explanations to users | user study 1 · ImageNet and CUB | user accuracy on model-incorrect predictions: 0.447 and 0.294 for incorrect predictions | No clear effect |
| show EPIC prototype explanations to users | user study 2 · ImageNet and CUB | user accuracy at identifying chosen class: 0.568 (ImageNet), 0.55 (CUB) vs chance | Improved |
| optimize disentanglement matrix for prototype purity (ablation) | pretrained ResNet · prototype quality (qualitative) | prototype coherence | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- PIPNet — PIP-Net: Patch-Based Intuitive Prototypes for Interpretable Image Classification (CVPR 2023) · better on 0 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- I-CEE: Tailoring Explanations of Image Classification Models to User Expertise (AAAI 2024 · similarity 0.88)
- Navigating the Maze of Explainable AI: A Systematic Approach to Evaluating Methods and Metrics (NeurIPS 2024 · similarity 0.86)
- Post-hoc Part-Prototype Networks (ICML 2024 · similarity 0.85)
- Keep the Faith: Faithful Explanations in Convolutional Neural Networks for Case-Based Reasoning (AAAI 2024 · similarity 0.85)
- How to Probe: Simple Yet Effective Techniques for Improving Post-hoc Explanations (ICLR 2025 · similarity 0.85)
- Probabilistic Conceptual Explainers: Trustworthy Conceptual Explanations for Vision Foundation Models (ICML 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
