Research map / Image generation / Fast diffusion sampling
GuidedBridge: Training-freely Improving Bridge Models with Prior Guidance
Zehua Chen, Yucheng Yang, Binjie Yuan, Kaiwen Zheng, Jun S. Liu, Jun Zhu
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 FMPG guidance to DBIM sampler | DBIM · Edges->Handbags NFE 20 | FID: 1.74 -> 1.07 | Improved |
| add FMPG guidance to DBIM sampler | DBIM · DIODE-Outdoor NFE 20 | FID: 4.99 -> 3.20 | Improved |
| add FMPG guidance to DBIM sampler | DBIM · DIODE-Outdoor NFE 100 | FID: 2.58 -> 2.06 | Improved |
| add PG to DDBM sampler | DDBM · Edges->Handbags NFE 300 | FID: 0.65 -> 0.59 | Improved |
| add PG to DDBM sampler | DDBM · Edges->Handbags NFE 150 | FID: 1.30 -> 1.23 | Improved |
| add frequency modulation to PG (noise) (ablation) | DBIM+PG · Edges->Handbags 10k NFE 10 | FID: 3.74 -> 3.51 | Improved |
| add frequency modulation to PG (blur) (ablation) | DBIM+PG · Edges->Handbags 10k NFE 10 | FID: 4.01 -> 3.75 | Improved |
| add frequency modulation to PG (noise) (ablation) | DBIM+PG · DIODE NFE 10 | FID: 6.25 -> 5.28 | Improved |
| add CFG before FMPG (CFG first, then FMPG) (ablation) | DBIM+FMPG · ImageNet inpainting NFE 10 | FID: 5.41 -> 3.86 | Improved |
| add FMPG after CFG (ablation) | DBIM+CFG · ImageNet inpainting NFE 10 | FID: 4.34 -> 3.86 | Improved |
| add CFG-FMPG to DBIM+CFG | DBIM+CFG · ImageNet inpainting NFE 10 | IS: 149.4 -> 154.4 | Improved |
| add CFG-FMPG to DBIM+CFG | DBIM+CFG · ImageNet inpainting NFE 40 | classification accuracy: 75.8 -> 75.6 | No clear effect |
Closest papers by meaning
- Adaptive Guidance: Training-free Acceleration of Conditional Diffusion Models (AAAI 2025 · similarity 0.89)
- Denoising Diffusion Bridge Models (ICLR 2024 · similarity 0.89)
- IRBridge: Solving Image Restoration Bridge with Pre-trained Generative Diffusion Models (ICML 2025 · similarity 0.88)
- Dynamic Negative Guidance of Diffusion Models (ICLR 2025 · similarity 0.88)
- Rectified Diffusion Guidance for Conditional Generation (CVPR 2025 · similarity 0.88)
- Eliminating Oversaturation and Artifacts of High Guidance Scales in Diffusion Models (ICLR 2025 · similarity 0.88)
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.
