Shifting the Breaking Point of Flow Matching for Multi-Instance Editing
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_ | 1866914317999276032 |
|---|---|
| author | Zaccagnino, Carmine Quattrini, Fabio Simsar, Enis Gazulla, Marta Tintoré Cucchiara, Rita Tonioni, Alessio Cascianelli, Silvia |
| author_facet | Zaccagnino, Carmine Quattrini, Fabio Simsar, Enis Gazulla, Marta Tintoré Cucchiara, Rita Tonioni, Alessio Cascianelli, Silvia |
| contents | Flow matching models have recently emerged as an efficient alternative to diffusion, especially for text-guided image generation and editing, offering faster inference through continuous-time dynamics. However, existing flow-based editors predominantly support global or single-instruction edits and struggle with multi-instance scenarios, where multiple parts of a reference input must be edited independently without semantic interference. We identify this limitation as a consequence of globally conditioned velocity fields and joint attention mechanisms, which entangle concurrent edits. To address this issue, we introduce Instance-Disentangled Attention, a mechanism that partitions joint attention operations, enforcing binding between instance-specific textual instructions and spatial regions during velocity field estimation. We evaluate our approach on both natural image editing and a newly introduced benchmark of text-dense infographics with region-level editing instructions. Experimental results demonstrate that our approach promotes edit disentanglement and locality while preserving global output coherence, enabling single-pass, instance-level editing. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_08749 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Shifting the Breaking Point of Flow Matching for Multi-Instance Editing Zaccagnino, Carmine Quattrini, Fabio Simsar, Enis Gazulla, Marta Tintoré Cucchiara, Rita Tonioni, Alessio Cascianelli, Silvia Computer Vision and Pattern Recognition Flow matching models have recently emerged as an efficient alternative to diffusion, especially for text-guided image generation and editing, offering faster inference through continuous-time dynamics. However, existing flow-based editors predominantly support global or single-instruction edits and struggle with multi-instance scenarios, where multiple parts of a reference input must be edited independently without semantic interference. We identify this limitation as a consequence of globally conditioned velocity fields and joint attention mechanisms, which entangle concurrent edits. To address this issue, we introduce Instance-Disentangled Attention, a mechanism that partitions joint attention operations, enforcing binding between instance-specific textual instructions and spatial regions during velocity field estimation. We evaluate our approach on both natural image editing and a newly introduced benchmark of text-dense infographics with region-level editing instructions. Experimental results demonstrate that our approach promotes edit disentanglement and locality while preserving global output coherence, enabling single-pass, instance-level editing. |
| title | Shifting the Breaking Point of Flow Matching for Multi-Instance Editing |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.08749 |