Research map / Label-efficient and robust learning / Recognition and long-tail
Towards Better Visualizing the Decision Basis of Networks via Unfold and Conquer Attribution Guidance
Jung-Ho Hong, Woo-Jeoung Nam, Kyu-Sung Jeon, Seong-Whan Lee
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 |
|---|---|---|---|
| Apply UCAG to GradCAM | GradCAM on ImageNet with ResNet50 · ImageNet | deletion AUC: 11.3 -> 8.6 | Improved |
| Apply UCAG to GradCAM | GradCAM on ImageNet with ResNet50 · ImageNet | insertion AUC: 53.9 -> 53.4 | No clear effect |
| Apply UCAG to GradCAM on VOC07 | GradCAM · VOC07 Test | pointing game: 86.6/74.0 -> 91.1/82.8 | Improved |
| Apply UCAG to GradCAM on COCO14 | GradCAM · COCO14 Val | pointing game: 54.2/49.0 -> 61.8/57.6 | Improved |
| Apply UCAG to GradCAM | GradCAM on ImageNet segmentation with ResNet50 · ImageNet segmentation | positive map density: 2.33 -> 3.89 | Improved |
| Apply UCAG to GradCAM | GradCAM on ImageNet segmentation with ResNet50 · ImageNet segmentation | negative map density: 0.86 -> 0.81 | Improved |
| Use image upscaling without spatial scrutiny (ablation) | GradCAM on ImageNet · ImageNet | deletion AUC | Hurt |
| Use image upscaling without spatial scrutiny (ablation) | GradCAM on ImageNet · ImageNet | insertion AUC | Hurt |
Closest papers by meaning
- Illuminating Salient Contributions in Neuron Activation With Attribution Equilibrium (TPAMI 2024 · similarity 0.89)
- How to Probe: Simple Yet Effective Techniques for Improving Post-hoc Explanations (ICLR 2025 · similarity 0.87)
- GOAt: Explaining Graph Neural Networks via Graph Output Attribution (ICLR 2024 · similarity 0.85)
- Path Choice Matters for Clear Attributions in Path Methods (ICLR 2024 · similarity 0.84)
- Keep the Faith: Faithful Explanations in Convolutional Neural Networks for Case-Based Reasoning (AAAI 2024 · similarity 0.84)
- Faithful and Accurate Self-Attention Attribution for Message Passing Neural Networks via the Computation Tree Viewpoint (AAAI 2025 · similarity 0.84)
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.
