Research map / Image generation / Fast diffusion sampling
Self-Guidance: Boosting Flow and Diffusion Generation on Their Own
Tiancheng Li, Weijian Luo, Zhiyang Chen, Liyuan Ma, Guo-Jun Qi
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 Self-Guidance alone | SD 1.4 · MS-COCO 2017 val | FID: 50.7602 -> 42.7362 | Improved |
| add SG to CFG | SD 1.4 · MS-COCO 2017 val | FID: 27.07 -> 26.67 | Improved |
| add SG-prev to CFG | SD 1.4 · MS-COCO 2017 val | FID: 27.07 -> 23.70 | Improved |
| add SG to CFG | SD 3 · MS-COCO 2017 val | FID: 27.00 -> 25.21 | Improved |
| add SG-prev to CFG | SD 3.5 · MS-COCO 2017 val | FID: 23.86 -> 20.74 | Improved |
| add SG to CFG | Flux.1 · MS-COCO 2017 val | FID: 29.74 -> 28.59 | Improved |
| add SG to CFG+PAG | SD 3 · MS-COCO 2017 val | FID: 27.34 -> 25.28 | Improved |
| add SG to CFG | SD 1.4 · HPS v2.1 | HPS v2.1 average: 24.24 -> 24.59 | Improved |
| add SG to CFG | SD 3 · MS-COCO 2017 val | Aesthetic Score: 5.3915 -> 5.4431 | Improved |
| add SG-prev to CFG | SD 3 · MS-COCO 2017 val | Aesthetic Score: 5.3915 -> 5.3260 | Hurt |
| add SG to CFG | Flux.1 · COCO hand prompts | FID-H: 67.0163 to 66.4360 | Improved |
| add SG to CogVideoX-5B | CogVideoX-5B · VBench | Human Action: 0.8320 -> 0.8450 | Improved |
Closest papers by meaning
- REG: Rectified Gradient Guidance for Conditional Diffusion Models (ICML 2025 · similarity 0.89)
- Plug-and-Play Diffusion Distillation (CVPR 2024 · similarity 0.89)
- Guiding a Diffusion Model with a Bad Version of Itself (NeurIPS 2024 · similarity 0.89)
- The Crystal Ball Hypothesis in diffusion models: Anticipating object positions from initial noise (ICLR 2025 · similarity 0.89)
- Shielded Diffusion: Generating Novel and Diverse Images using Sparse Repellency (ICML 2025 · similarity 0.89)
- Applying Guidance in a Limited Interval Improves Sample and Distribution Quality in Diffusion Models (NeurIPS 2024 · similarity 0.89)
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.
