From Diffusion to One-Step Generation: A Comparative Study of Flow-Based Models with Application to Image Inpainting
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915638919823360 |
|---|---|
| author | Agarwal, Umang Sangore, Rudraksh Laddha, Sumit |
| author_facet | Agarwal, Umang Sangore, Rudraksh Laddha, Sumit |
| contents | We present a comprehensive comparative study of three generative modeling paradigms: Denoising Diffusion Probabilistic Models (DDPM), Conditional Flow Matching (CFM), and MeanFlow. While DDPM and CFM require iterative sampling, MeanFlow enables direct one-step generation by modeling the average velocity over time intervals. We implement all three methods using a unified TinyUNet architecture (<1.5M parameters) on CIFAR-10, demonstrating that CFM achieves an FID of 24.15 with 50 steps, significantly outperforming DDPM (FID 402.98). MeanFlow achieves FID 29.15 with single-step sampling -- a 50X reduction in inference time. We further extend CFM to image inpainting, implementing mask-guided sampling with four mask types (center, random bbox, irregular, half). Our fine-tuned inpainting model achieves substantial improvements: PSNR increases from 4.95 to 8.57 dB on center masks (+73%), and SSIM improves from 0.289 to 0.418 (+45%), demonstrating the effectiveness of inpainting-aware training. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_21215 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | From Diffusion to One-Step Generation: A Comparative Study of Flow-Based Models with Application to Image Inpainting Agarwal, Umang Sangore, Rudraksh Laddha, Sumit Computer Vision and Pattern Recognition Machine Learning We present a comprehensive comparative study of three generative modeling paradigms: Denoising Diffusion Probabilistic Models (DDPM), Conditional Flow Matching (CFM), and MeanFlow. While DDPM and CFM require iterative sampling, MeanFlow enables direct one-step generation by modeling the average velocity over time intervals. We implement all three methods using a unified TinyUNet architecture (<1.5M parameters) on CIFAR-10, demonstrating that CFM achieves an FID of 24.15 with 50 steps, significantly outperforming DDPM (FID 402.98). MeanFlow achieves FID 29.15 with single-step sampling -- a 50X reduction in inference time. We further extend CFM to image inpainting, implementing mask-guided sampling with four mask types (center, random bbox, irregular, half). Our fine-tuned inpainting model achieves substantial improvements: PSNR increases from 4.95 to 8.57 dB on center masks (+73%), and SSIM improves from 0.289 to 0.418 (+45%), demonstrating the effectiveness of inpainting-aware training. |
| title | From Diffusion to One-Step Generation: A Comparative Study of Flow-Based Models with Application to Image Inpainting |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2511.21215 |