Delta Rectified Flow Sampling for Text-to-Image Editing
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866911585745764352 |
|---|---|
| author | Beaudouin, Gaspard Li, Minghan Kim, Jaeyeon Yoon, Sung-Hoon Wang, Mengyu |
| author_facet | Beaudouin, Gaspard Li, Minghan Kim, Jaeyeon Yoon, Sung-Hoon Wang, Mengyu |
| contents | We propose Delta Rectified Flow Sampling (DRFS), a novel inversion-free, path-aware editing framework within rectified flow models for text-to-image editing. DRFS is a distillation-based method that explicitly models the discrepancy between the source and target velocity fields in order to mitigate over-smoothing artifacts rampant in prior distillation sampling approaches. We further introduce a time-dependent shift term to push noisy latents closer to the target trajectory, enhancing the alignment with the target distribution. We theoretically demonstrate that disabling this shift recovers Delta Denoising Score (DDS), bridging score-based diffusion optimization and velocity-based rectified-flow optimization. Moreover, under rectified-flow dynamics, a linear shift schedule recovers the inversion-free method FlowEdit as a strict special case, yielding a unifying view of optimization and ODE editing. We conduct an analysis to guide the design of our shift term, and experimental results on the widely used PIE Benchmark indicate that DRFS achieves superior editing quality, fidelity, and controllability while requiring no architectural modifications. Code is available at https://github.com/Harvard-AI-and-Robotics-Lab/DeltaRectifiedFlowSampling. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_05342 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Delta Rectified Flow Sampling for Text-to-Image Editing Beaudouin, Gaspard Li, Minghan Kim, Jaeyeon Yoon, Sung-Hoon Wang, Mengyu Computer Vision and Pattern Recognition Machine Learning We propose Delta Rectified Flow Sampling (DRFS), a novel inversion-free, path-aware editing framework within rectified flow models for text-to-image editing. DRFS is a distillation-based method that explicitly models the discrepancy between the source and target velocity fields in order to mitigate over-smoothing artifacts rampant in prior distillation sampling approaches. We further introduce a time-dependent shift term to push noisy latents closer to the target trajectory, enhancing the alignment with the target distribution. We theoretically demonstrate that disabling this shift recovers Delta Denoising Score (DDS), bridging score-based diffusion optimization and velocity-based rectified-flow optimization. Moreover, under rectified-flow dynamics, a linear shift schedule recovers the inversion-free method FlowEdit as a strict special case, yielding a unifying view of optimization and ODE editing. We conduct an analysis to guide the design of our shift term, and experimental results on the widely used PIE Benchmark indicate that DRFS achieves superior editing quality, fidelity, and controllability while requiring no architectural modifications. Code is available at https://github.com/Harvard-AI-and-Robotics-Lab/DeltaRectifiedFlowSampling. |
| title | Delta Rectified Flow Sampling for Text-to-Image Editing |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2509.05342 |