Dual-Solver: A Generalized ODE Solver for Diffusion Models with Dual Prediction
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910040485527552 |
|---|---|
| author | Park, Soochul Lee, Yeon Ju |
| author_facet | Park, Soochul Lee, Yeon Ju |
| contents | Diffusion models achieve state-of-the-art image quality. However, sampling is costly at inference time because it requires a large number of function evaluations (NFEs). To reduce NFEs, classical ODE numerical methods have been adopted. Yet, the choice of prediction type and integration domain leads to different sampling behaviors. To address these issues, we introduce Dual-Solver, which generalizes multistep samplers through learnable parameters that continuously (i) interpolate among prediction types, (ii) select the integration domain, and (iii) adjust the residual terms. It retains the standard predictor-corrector structure while preserving second-order local accuracy. These parameters are learned via a classification-based objective using a frozen pretrained classifier (e.g., MobileNet or CLIP). For ImageNet class-conditional generation (DiT, GM-DiT) and text-to-image generation (SANA, PixArt-$α$), Dual-Solver improves FID and CLIP scores in the low-NFE regime ($3 \le$ NFE $\le 9$) across backbones. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_03973 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Dual-Solver: A Generalized ODE Solver for Diffusion Models with Dual Prediction Park, Soochul Lee, Yeon Ju Machine Learning Computer Vision and Pattern Recognition 65L05, 68T07, 65C20 I.2.6; I.4.8; G.1.7 Diffusion models achieve state-of-the-art image quality. However, sampling is costly at inference time because it requires a large number of function evaluations (NFEs). To reduce NFEs, classical ODE numerical methods have been adopted. Yet, the choice of prediction type and integration domain leads to different sampling behaviors. To address these issues, we introduce Dual-Solver, which generalizes multistep samplers through learnable parameters that continuously (i) interpolate among prediction types, (ii) select the integration domain, and (iii) adjust the residual terms. It retains the standard predictor-corrector structure while preserving second-order local accuracy. These parameters are learned via a classification-based objective using a frozen pretrained classifier (e.g., MobileNet or CLIP). For ImageNet class-conditional generation (DiT, GM-DiT) and text-to-image generation (SANA, PixArt-$α$), Dual-Solver improves FID and CLIP scores in the low-NFE regime ($3 \le$ NFE $\le 9$) across backbones. |
| title | Dual-Solver: A Generalized ODE Solver for Diffusion Models with Dual Prediction |
| topic | Machine Learning Computer Vision and Pattern Recognition 65L05, 68T07, 65C20 I.2.6; I.4.8; G.1.7 |
| url | https://arxiv.org/abs/2603.03973 |