Research map / Label-efficient and robust learning / Recognition and long-tail
This Looks Like That: Deep Learning for Interpretable Image Recognition
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use ProtoPNet instead of the non-interpretable baseline | VGG16 · CUB-200-2011 cropped images | accuracy: 74.6 ± 0.2 -> 76.1 ± 0.2 | Helped |
| Use ProtoPNet instead of the non-interpretable baseline | VGG19 · CUB-200-2011 cropped images | accuracy: 75.1 ± 0.4 -> 78.0 ± 0.2 | Helped |
| Use ProtoPNet instead of the non-interpretable baseline | ResNet-34 · CUB-200-2011 cropped images | accuracy: 82.3 ± 0.3 -> 79.2 ± 0.1 | Hurt |
| Use ProtoPNet instead of the non-interpretable baseline | ResNet-152 · CUB-200-2011 cropped images | accuracy: 81.5 ± 0.4 -> 78.0 ± 0.3 | Hurt |
| Use ProtoPNet instead of the non-interpretable baseline | DenseNet-121 · CUB-200-2011 cropped images | accuracy: 80.5 ± 0.1 -> 80.2 ± 0.2 | No clear effect |
| Use ProtoPNet instead of the non-interpretable baseline | DenseNet-161 · CUB-200-2011 cropped images | accuracy: 82.2 ± 0.2 -> 80.1 ± 0.3 | Hurt |
| Combine logits from several ProtoPNet models | ProtoPNet · CUB-200-2011 cropped images | accuracy: 84.8% | Helped |
| Combine logits from several ProtoPNet models | ProtoPNet · CUB-200-2011 full images | accuracy: 72.7%, 74.4%, and 75.7% -> 80.8% | Helped |
| Combine VGG19-, ResNet-34-, and DenseNet-121-based ProtoPNets | ProtoPNet · Stanford Cars | accuracy: 91.4% | Helped |
Closest papers by meaning
- Interpretable Image Classification via Non-parametric Part Prototype Learning (CVPR 2025 · similarity 0.93)
- ProtoArgNet: Interpretable Image Classification with Super-Prototypes and Argumentation (AAAI 2025 · similarity 0.91)
- LucidPPN: Unambiguous Prototypical Parts Network for User-centric Interpretable Computer Vision (ICLR 2025 · similarity 0.91)
- Improving Prototypical Visual Explanations with Reward Reweighing, Reselection, and Retraining (ICML 2024 · similarity 0.90)
- Interpretable Image Classification with Adaptive Prototype-based Vision Transformers (NeurIPS 2024 · similarity 0.90)
- Rashomon Sets for Prototypical-Part Networks: Editing Interpretable Models in Real-Time (CVPR 2025 · similarity 0.89)
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.