Research map / Detection and segmentation / CNN features
Harmonized Feature Conditioning and Frequency-Prompt Personalization for Multi-Rater Medical Segmentation
Sanaz Karimijafarbigloo, Armin Khosravi, Alireza Kheyrkhah, Reza Azad, Mauricio Reyes, Dorit Merhof
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 |
|---|---|---|---|
| increase number of prior samples K from 10 to 50 | Harmonizer Network · LIDC-IDRI | GED: 0.1164 -> 0.1048 | Improved |
| increase number of prior samples K from 10 to 50 | Harmonizer Network · NPC-170 | GED: 0.1960 -> 0.1758 | Improved |
| increase number of prior samples K from 10 to 50 | Prob. U-Net · LIDC-IDRI | GED: 0.2181 -> 0.2168 | No clear effect |
| add Noise Harmonizer and GED constraint (phase I) versus D-Persona prior | Harmonizer Network · LIDC-IDRI | GED: 0.105 vs. 0.136 | Improved |
| add Noise Harmonizer and GED constraint (phase I) versus D-Persona prior | Harmonizer Network · NPC-170 | GED: 0.175 vs. 0.197 | Improved |
| add frequency-domain prompts (phase II) on top of harmonized features, versus D-Persona | Harmonizer Network · LIDC-IDRI | Dice_mean: 89.17 -> 90.78 | Improved |
| add frequency-domain prompts versus D-Persona | Harmonizer Network · NPC-170 | Dice_mean: 80.4 -> 81.63 | Improved |
| add GED regularizer during training (ablation) | Harmonizer Network · LIDC-IDRI | GED vs K | Improved |
| add frequency adapter to the decoder (ablation) | Harmonizer Network · LIDC-IDRI | high-frequency spectral energy | Improved |
| add harmonization under Gaussian noise sigma=0.25 | Harmonizer Network vs D-Persona · LIDC-IDRI | DSC: 71.11 -> 84.27 | Improved |
| add harmonization under gamma jitter 1.75 | Harmonizer Network vs D-Persona · LIDC-IDRI | DSC: 81.53 -> 86.23 | Improved |
| vary lesion size bin (small, medium, large) (ablation) | Harmonizer Network · LIDC-IDRI | Dice | No clear effect |
Closest papers by meaning
- Diversified and Personalized Multi-rater Medical Image Segmentation (CVPR 2024 · similarity 0.90)
- Training Like a Medical Resident: Context-Prior Learning Toward Universal Medical Image Segmentation (CVPR 2024 · similarity 0.87)
- Versatile Medical Image Segmentation Learned from Multi-Source Datasets via Model Self-Disambiguation (CVPR 2024 · similarity 0.86)
- S2S2: Semantic Stacking for Robust Semantic Segmentation in Medical Imaging (AAAI 2025 · similarity 0.86)
- GapMatch: Bridging Instance and Model Perturbations for Enhanced Semi-Supervised Medical Image Segmentation (AAAI 2025 · similarity 0.86)
- Enhancing SAM with Efficient Prompting and Preference Optimization for Semi-supervised Medical Image Segmentation (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.
