Research map / Image generation / Fast diffusion sampling
Set-Coupled Guidance: Set-Level Coordination in Diffusion-Based Dataset Distillation
Ziang Gan, Qi Zhu, Libao Zhang
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 SCG set-level guidance | DiT · ImageNette IPC=10, ResNet-18 | top-1 accuracy: 58.2 -> 65.8 | Improved |
| add SCG | MGD3 · ImageNette IPC=50, ResNet-18 | top-1 accuracy: 78.1 -> 82.6 | Improved |
| add SCG | CaO2 · ImageNette IPC=10, ResNet-18 | top-1 accuracy: 61.5 -> 62.8 | Improved |
| add SCG | MGD3 · ImageWoof IPC=50, ResNet-18 | top-1 accuracy: 58.3 -> 58.0 | No clear effect |
| add SCG | MGD3 · ImageNet-1K IPC=10, ResNet-18 | top-1 accuracy: 21.2 -> 22.9 | Improved |
| add kernel coupling (alpha) alone (ablation) | MGD3+SCG · ImageNette IPC=10, ResNet-18 | top-1 accuracy: 61.5 -> 62.2 | Improved |
| add spectral regulation (beta) alone (ablation) | MGD3+SCG · ImageNette IPC=10, ResNet-18 | top-1 accuracy: 61.5 -> 63.4 | Improved |
| add coupling on top of regulation (ablation) | MGD3+SCG · ImageNette IPC=10, ResNet-18 | top-1 accuracy: 63.4 -> 64.1 | Improved |
| add kernel coupling alone (ablation) | MGD3+SCG · ImageNette IPC=10, ResNetAP-10 | top-1 accuracy: 65.1 -> 64.2 | Hurt |
| add SCG controller to batched sampling | MGD3 · ImageNette IPC=10 | sampling time (s): 8.20 -> 8.35 | No clear effect |
| batch IPC samples per forward pass instead of sequential | CaO2 · ImageNette IPC=10 | sampling time (s): 56.79 -> 16.11 | Improved |
| vary controller weights (alpha, beta) (ablation) | SCG · ImageNette IPC=10, ResNet-18 | top-1 accuracy | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- CaO2 — CaO2: Rectifying Inconsistencies in Diffusion-Based Dataset Distillation (ICCV 2025) · better on 4 of 4 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
- Efficient Dataset Distillation via Minimax Diffusion (CVPR 2024 · similarity 0.87)
- Feedback Guidance of Diffusion Models (NeurIPS 2025 · similarity 0.86)
- MGD$^3$ : Mode-Guided Dataset Distillation using Diffusion Models (ICML 2025 · similarity 0.86)
- Plug-and-Play Diffusion Distillation (CVPR 2024 · similarity 0.86)
- Understanding and Improving Training-free Loss-based Diffusion Guidance (NeurIPS 2024 · similarity 0.86)
- Eliminating Oversaturation and Artifacts of High Guidance Scales in Diffusion Models (ICLR 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.
