Research map / Image generation / Text-to-image diffusion
Making Training-Free Diffusion Segmentors Scale with the Generative Power
Benyuan Meng, Qianqian Xu, Zitai Wang, Xiaochun Cao, Longtao Huang, Qingming Huang
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 |
|---|---|---|---|
| add auto head-wise aggregation (ablation) | GoCA on SD XL · Pascal VOC 2012 | mIoU: 51.1 -> 56.1 | Improved |
| add auto head-wise aggregation (ablation) | GoCA on SD v1.5 · Pascal VOC 2012 | mIoU: 44.3 -> 44.8 | No clear effect |
| add auto layer-wise aggregation (ablation) | GoCA on SD v1.5 · Pascal VOC 2012 | mIoU: 44.3 -> 52.1 | Improved |
| add auto layer-wise aggregation (ablation) | GoCA on SD XL · Pascal VOC 2012 | mIoU: 51.1 -> 51.3 | No clear effect |
| add per-pixel rescaling (ablation) | GoCA on SD v1.5 · Pascal VOC 2012 | mIoU: 44.3 -> 52.6 | Improved |
| add per-pixel rescaling (ablation) | GoCA on SD XL · Pascal VOC 2012 | mIoU: 51.1 -> 51.4 | No clear effect |
| combine head, layer and rescaling (ablation) | GoCA on SD XL · Pascal VOC 2012 | mIoU: 51.1 -> 65.6 | Improved |
| combine all components on SD v1.5 (ablation) | GoCA · Pascal VOC 2012 | mIoU: 44.3 -> 60.7 | Improved |
| swap SD v1.5 for Flux with vanilla averaging | Vanilla · Pascal-Context-59 | mIoU: 32.3 -> 48.4 | Improved |
| use MSE layer metric instead of dot-product (ablation) | GoCA on SD v1.5 · Pascal-Context-59 | mIoU: 40.4 -> 33.9 | Hurt |
| add GoCA to S-CFG | S-CFG on SD v1.5 · COCO-30k CFG=2.0 | FID: 36.45 -> 35.23 | Improved |
| add GoCA to S-CFG, three seeds | S-CFG on SD v1.5 · COCO-30k CFG=5.0 | FID: 19.36 -> 19.05 | Improved |
Closest papers by meaning
- Seg4Diff: Unveiling Open-Vocabulary Semantic Segmentation in Text-to-Image Diffusion Transformers (NeurIPS 2025 · similarity 0.91)
- EmerDiff: Emerging Pixel-level Semantic Knowledge in Diffusion Models (ICLR 2024 · similarity 0.90)
- Open-Vocabulary Attention Maps with Token Optimization for Semantic Segmentation in Diffusion Models (CVPR 2024 · similarity 0.90)
- Text-Image Alignment for Diffusion-Based Perception (CVPR 2024 · similarity 0.90)
- Repurposing Stable Diffusion Attention for Training-Free Unsupervised Interactive Segmentation (CVPR 2025 · similarity 0.90)
- Unleashing the Potential of the Diffusion Model in Few-shot Semantic Segmentation (NeurIPS 2024 · similarity 0.90)
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.
