Edit-R2: Context-Aware Reinforcement Learning for Multi-Turn Image Editing

• URL: http://arxiv.org/abs/2606.05950v1
• Date: Thu, 04 Jun 2026 09:49:47 GMT
• Title: Edit-R2: Context-Aware Reinforcement Learning for Multi-Turn Image Editing
• Abstract summary: We introduce Edit-R2, a novel reinforcement learning framework for unified multimodal models.<n>It reconstructs operative session intent, which effectively consolidates scattered historical constraints into an explicit reasoning trace before each editing turn.<n>It achieves competitive performance compared against strong baselines.
• Score: 42.176441824728066
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Text-guided image editing has advanced rapidly with diffusion models and unified multimodal foundation models. However, most existing methods remain confined to single-turn settings, overlooking the more realistic scenario of multi-turn in-context editing, where users iteratively refine an image through a sequence of instructions. In this setting, a model must follow each new instruction while preserving accumulated session-level constraints, challenged by two coupled failure modes: long-context dilution, where sparse textual constraints become difficult to recover from growing interleaved image-text histories, and state contamination, where earlier editing mistakes degrade subsequent generations. We introduce Edit-R2, a novel reinforcement learning post-training framework for unified multimodal models. Edit-R2 reconstructs the operative session intent, which effectively consolidates scattered historical constraints into an explicit reasoning trace before each editing turn. It further enables multi-turn RL over both reasoning and generation through a unified objective that jointly optimizes intent reconstruction generation in discrete text space and flow-matching image generation in continuous latent space, while a trajectory filtering mechanism suppresses corrupted rollouts to stabilize training under state contamination. To support systematic evaluation, we introduce MICE-Bench, a large-scale benchmark for multi-turn in-context editing with automated metrics for instruction following (IF), content consistency (CC), and global awareness (GA) over accumulated session constraints. Experiments show that Edit-R2 substantially improves multi-turn in-context editing and achieves competitive performance compared against strong baselines.

Related papers

• ContextMaster: Interactive Multi-Shot Video Creation via Fixed-Budget Sparse Context Routing [42.25675968821303]
  Video models increasingly support generation, reference conditioning, and editing within a single model.<n>We formalize this setting as interactive multi-shot video creation (IMVC)<n>We introduce ContextMaster, a unified model with a role-aware context representation for these operations.
  arXiv  Detail & Related papers  (2026-08-05T15:27:26Z)
• AnchorEdit: Maintaining Temporal Consistency in Multi-turn Image Editing via Causal Memory [58.322260363577016]
  Multi-turn image editing is essential for iterative design, yet current models often struggle with identity drift and error accumulation over successive steps.<n>We propose AnchorEdit, the first autoregressive (AR) diffusion-based framework designed specifically for high-resolution, long-term multi-turn editing.
  arXiv  Detail & Related papers  (2026-06-10T07:26:31Z)
• Multimodal Large Language Models for Multi-Subject In-Context Image Generation [56.20395856287325]
  Music is the first MLLM specifically designed for textbfMUlti-textbfSubject textbfIn-textbfContext image generation.<n>We enhance the model's understanding of multi-subject semantic relationships through a vision chain-of-thought mechanism.<n>By incorporating complex subject images during training, we improve the model's capacity for chained reasoning.
  arXiv  Detail & Related papers  (2026-04-08T15:37:42Z)
• ITO: Images and Texts as One via Synergizing Multiple Alignment and Training-Time Fusion [16.28499377328499]
  ITO is a framework addressing the limitation through two synergistic mechanisms.<n>We show that ITO consistently outperforms strong baselines across classification, retrieval, and multimodal benchmarks.<n>Our analysis reveals that while multiple alignment drives discriminative power, training-time fusion acts as a critical structural regularizer.
  arXiv  Detail & Related papers  (2026-03-03T09:08:53Z)
• UniRef-Image-Edit: Towards Scalable and Consistent Multi-Reference Image Editing [33.64590153603506]
  We present UniRef-Image-Edit, a high-performance multi-modal generation system.<n>It unifies single-image editing and multi-image composition within a single framework.
  arXiv  Detail & Related papers  (2026-02-15T15:24:03Z)
• CRAFT: Continuous Reasoning and Agentic Feedback Tuning for Multimodal Text-to-Image Generation [0.0]
  CRAFT (Continuous Reasoning and Agentic Feedback Tuning) is a training-free, model-agnostic framework that brings structured reasoning paradigm to multimodal image generation.<n>It consistently improves compositional accuracy, text rendering, and preference-based evaluations.<n>These improvements incur only a negligible inference-time overhead, allowing smaller or cheaper models to approach the quality of substantially more expensive systems.
  arXiv  Detail & Related papers  (2025-12-23T13:44:41Z)
• IAR2: Improving Autoregressive Visual Generation with Semantic-Detail Associated Token Prediction [77.06211178777939]
  IAR2 is an advanced autoregressive framework that enables a hierarchical semantic-detail synthesis process.<n>We show that IAR2 sets a new state-of-the-art for autoregressive image generation, achieving a FID of 1.50 on ImageNet.
  arXiv  Detail & Related papers  (2025-10-08T12:08:21Z)
• CycleVAR: Repurposing Autoregressive Model for Unsupervised One-Step Image Translation [9.628074306577851]
  Current conditional autoregressive image generation methods have shown promising results, yet their potential remains largely unexplored in the practical unsupervised image translation domain.<n>A critical limitation stems from the discrete quantization inherent in traditional Vector Quantization-based frameworks.<n>We propose Softmax Relaxed Quantization, a novel approach that reformulates codebook selection as a continuous probability mixing process.
  arXiv  Detail & Related papers  (2025-06-29T17:43:04Z)
• DCI: Dual-Conditional Inversion for Boosting Diffusion-Based Image Editing [73.12011187146481]
  Inversion within Diffusion models aims to recover the latent noise representation for a real or generated image.<n>Most inversion approaches suffer from an intrinsic trade-off between reconstruction accuracy and editing flexibility.<n>We introduce Dual-Conditional Inversion (DCI), a novel framework that jointly conditions on the source prompt and reference image.
  arXiv  Detail & Related papers  (2025-06-03T07:46:44Z)
• Multimodal LLM-Guided Semantic Correction in Text-to-Image Diffusion [52.315729095824906]
  MLLM Semantic-Corrected Ping-Pong-Ahead Diffusion (PPAD) is a novel framework that introduces a Multimodal Large Language Model (MLLM) as a semantic observer during inference.<n>It performs real-time analysis on intermediate generations, identifies latent semantic inconsistencies, and translates feedback into controllable signals that actively guide the remaining denoising steps.<n>Extensive experiments demonstrate PPAD's significant improvements.
  arXiv  Detail & Related papers  (2025-05-26T14:42:35Z)
• Uniform Attention Maps: Boosting Image Fidelity in Reconstruction and Editing [66.48853049746123]
  We analyze reconstruction from a structural perspective and propose a novel approach that replaces traditional cross-attention with uniform attention maps.<n>Our method effectively minimizes distortions caused by varying text conditions during noise prediction.<n> Experimental results demonstrate that our approach not only excels in achieving high-fidelity image reconstruction but also performs robustly in real image composition and editing scenarios.
  arXiv  Detail & Related papers  (2024-11-29T12:11:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.