Research map / Image generation / Fast diffusion sampling
Accelerated Likelihood Maximization for Diffusion-based Versatile Content Generation
Hyunsoo Lee, Inwoo Hwang, Young Min Kim
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 the likelihood maximization over the unobserved region (ablation) | ALM · AFHQ inpainting | M-SSIM: 0.300 -> 0.351 | Improved |
| add the likelihood maximization over the unobserved region (ablation) | ALM · CelebA-HQ inpainting | M-SSIM: 0.332 -> 0.417 | Improved |
| add the conditional likelihood term (ablation) | ALM · AFHQ inpainting | M-SSIM: 0.323 -> 0.351 | Improved |
| add the conditional likelihood term (ablation) | ALM · CelebA-HQ inpainting | M-SSIM: 0.357 -> 0.417 | Improved |
| add the joint log-density term (ablation) | ALM · AFHQ inpainting | M-SSIM: 0.327 -> 0.351 | Improved |
| add the joint log-density term (ablation) | ALM · CelebA-HQ inpainting | M-SSIM: 0.385 -> 0.417 | Improved |
| use one-step approximation instead of iterative updates (ablation) | ALM · AFHQ inpainting | M-SSIM: 0.298 -> 0.351 | No clear effect |
| use one-step approximation instead of iterative updates (ablation) | ALM · CelebA-HQ inpainting | LPIPS: 0.277 -> 0.251 | No clear effect |
| ALM vs SyncSDE on inpainting | ALM · AFHQ inpainting | M-SSIM: 0.302 -> 0.351 | Improved |
| ALM vs SDI on AFHQ MSE | ALM · AFHQ inpainting | MSE: 0.140 -> 0.143 | Hurt |
| ALM vs SyncSDE on wide image generation | ALM · Wide image generation | KID: 51.84 -> 42.98 | Improved |
| swap backbone from SD to FLUX | ALM · AFHQ inpainting | M-SSIM: 0.351 -> 0.277 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- HD-Painter — HD-Painter: High-Resolution and Prompt-Faithful Text-Guided Image Inpainting with Diffusion Models (ICLR 2025) · better on 2 of 2 shared results
- StochSync — StochSync: Stochastic Diffusion Synchronization for Image Generation in Arbitrary Spaces (ICLR 2025) · better on 2 of 2 shared results
- BrushNet — BrushNet: A Plug-and-Play Image Inpainting Model with Decomposed Dual-Branch Diffusion (ECCV 2024) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Diffusion Tree Sampling: Scalable inference‑time alignment of diffusion models (NeurIPS 2025 · similarity 0.89)
- Accelerating Convergence of Score-Based Diffusion Models, Provably (ICML 2024 · similarity 0.89)
- Test-time Alignment of Diffusion Models without Reward Over-optimization (ICLR 2025 · similarity 0.89)
- Adaptive Non-Uniform Timestep Sampling for Accelerating Diffusion Model Training (CVPR 2025 · similarity 0.89)
- The Emergence of Reproducibility and Consistency in Diffusion Models (ICML 2024 · similarity 0.88)
- Elucidating the Exposure Bias in Diffusion Models (ICLR 2024 · 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.
