Shifting the Breaking Point of Flow Matching for Multi-Instance Editing

2026-02-09

AI Summary

This work addresses the challenge of achieving independent, non-interfering simultaneous editing of multiple regions in multi-instance scenarios—a limitation of existing flow-matching-based image editing methods. To this end, the authors propose an instance-decoupled attention mechanism that explicitly binds textual instructions to their corresponding spatial regions within the flow-matching framework, enabling locally controllable multi-instance editing in a single forward pass. By circumventing the constraints imposed by global conditional velocity fields and joint attention mechanisms on edit disentanglement, the method achieves high-fidelity, spatially precise, and globally consistent edits. Its effectiveness is validated on both natural images and a newly introduced benchmark of text-dense infographics, demonstrating superior performance in complex, multi-object editing tasks.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

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.

Problem

• flow matching
• multi-instance editing
• image editing
• semantic interference
• instance disentanglement

Innovation

• Flow Matching
• Multi-Instance Editing
• Instance-Disentangled Attention
• Text-Guided Image Editing
• Velocity Field