Research map / Label-efficient and robust learning / Recognition and long-tail
Cartoon Explanations of Image Classifiers
Stefan Kolek, Duc Anh Nguyen, Ron Levie, Joan Bruna, Gitta Kutyniok
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 |
|---|---|---|---|
| use Gaussian adaptive noise instead of constant zero perturbations as obfuscation in CartoonX (ablation) | CartoonX · ImageNet samples (qualitative) | explanation quality | Improved |
| vary the distortion measure (squared prob, l2, KL, distance to 1) (ablation) | CartoonX · ImageNet samples (qualitative) | explanation quality | No clear effect |
| use Haar instead of other mother wavelets (ablation) | CartoonX · ImageNet samples (qualitative) | explanation quality | Hurt |
| compute RDE in the wavelet domain (CartoonX) instead of pixel domain (Pixel RDE) | RDE · ImageNet (100 random samples) | rate-distortion / sparsity | Improved |
| use CartoonX instead of Pixel RDE for sparsity level sensitivity | RDE · ImageNet samples (qualitative) | sensitivity to lambda | Improved |
| use CartoonX instead of Pixel RDE | RDE · MobileNetV3-Small, ImageNet | runtime per image (s): 34.09s -> 45.10s | Hurt |
Closest papers by meaning
- Reproducibility Study of ”CartoonX: Cartoon Explanations of Image Classifiers” (NeurIPS 2023 · similarity 0.91)
- Explaining Image Classifiers With Multiscale Directional Image Representation (CVPR 2023 · similarity 0.89)
- Explanations for Occluded Images (ICCV 2021 · similarity 0.86)
- Explaining in Diffusion: Explaining a Classifier with Diffusion Semantics (CVPR 2025 · similarity 0.84)
- From Pixels to Perception: Interpretable Predictions via Instance-wise Grouped Feature Selection (ICML 2025 · similarity 0.82)
- SLICE: Stabilized LIME for Consistent Explanations for Image Classification (CVPR 2024 · similarity 0.82)
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.
