Research map / Architectures and efficiency / CNN design
LFI-CAM: Learning Feature Importance for Better Visual Explanation
Kwang Hee Lee, Chaewon Park, Junghyun Oh, Nojun Kwak
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 |
|---|---|---|---|
| replace ABN attention branch with LFI-CAM Feature Importance Network | ResNet18 · STL10 | top-1 error: 18.75 -> 18.16 | Improved |
| replace ABN attention branch with LFI-CAM Feature Importance Network | ResNet18 · Cat&Dog | top-1 error: 3.07 -> 2.72 | Improved |
| replace ABN attention branch with LFI-CAM Feature Importance Network | ResNet110 · CIFAR10 | top-1 error: 4.91 -> 5.73 | Hurt |
| replace ABN attention branch with LFI-CAM Feature Importance Network | ResNet110 · CIFAR100 | top-1 error: 22.82 -> 23.33 | Hurt |
| replace ABN attention branch with LFI-CAM Feature Importance Network | ResNeXt · CIFAR10 | top-1 error: 3.8 -> 4.27 | Hurt |
| replace ABN attention branch with LFI-CAM Feature Importance Network | ResNeXt · CIFAR100 | top-1 error: 17.7 -> 18.23 | Hurt |
| replace ABN attention branch with LFI-CAM Feature Importance Network | ResNet18 · ImageNet val | top-1 error: 28.98 -> 27.75 | Improved |
| replace ABN attention branch with LFI-CAM Feature Importance Network | ResNet152 · ImageNet val | top-1 error: 21.4 -> 21.95 | Hurt |
| add LFI-CAM to ResNet110 | ResNet110 · CIFAR10 | top-1 error: 6.43 -> 5.73 | Improved |
| add LFI-CAM to ResNeXt | ResNeXt · CIFAR10 | top-1 error: 3.84 -> 4.27 | Hurt |
| add LFI-CAM to ResNet18 | ResNet18 · ImageNet val | top-1 error: 30.24 -> 27.75 | Improved |
| add LFI-CAM to ResNet152 | ResNet152 · ImageNet val | top-1 error: 21.69 -> 21.95 | Hurt |
Closest papers by meaning
- Towards Better Explanations of Class Activation Mapping (ICCV 2021 · similarity 0.90)
- Empowering CAM-Based Methods with Capability to Generate Fine-Grained and High-Faithfulness Explanations (AAAI 2024 · similarity 0.88)
- Finer-CAM: Spotting the Difference Reveals Finer Details for Visual Explanation (CVPR 2025 · similarity 0.87)
- DD-CAM: Minimal Sufficient Explanations for Vision Models Using Delta Debugging (NeurIPS 2026 · similarity 0.87)
- CAPE: CAM as a Probabilistic Ensemble for Enhanced DNN Interpretation (CVPR 2024 · similarity 0.86)
- DiffCAM: Data-Driven Saliency Maps by Capturing Feature Differences (CVPR 2025 · similarity 0.86)
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.
