AnchorEdit: Maintaining Temporal Consistency in Multi-turn Image Editing via Causal Memory

• URL: http://arxiv.org/abs/2606.11751v2
• Date: Mon, 15 Jun 2026 08:11:46 GMT
• Title: AnchorEdit: Maintaining Temporal Consistency in Multi-turn Image Editing via Causal Memory
• Abstract summary: 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.
• Score: 58.322260363577016
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Multi-turn image editing is essential for iterative design, yet current models often struggle with identity drift and error accumulation over successive steps. While existing research leverages video priors for consistency, their reliance on bidirectional attention is fundamentally misaligned with the causal, sequential nature of interactive editing. In this paper, we propose AnchorEdit, the first autoregressive (AR) diffusion-based framework designed specifically for high-resolution, long-term multi-turn editing. AnchorEdit bridges the gap between video priors and causal inference through a three-stage training curriculum: identity-preserving sing-turn pretraining, causal AR forcing fine-tuning with a novel self-rollout strategy to mitigate exposure bias, and consistency distillation for efficient 4-step generation. During inference, we introduce a memory mechanism to anchor the initial subject identity and ensure stable extrapolation across extended editing trajectories. To evaluate performance, we provide a new high-resolution multi-turn editing benchmark designed to stress-test long-horizon stability. Extensive experiments demonstrate that AnchorEdit achieves state-of-the-art results, maintaining exceptional subject fidelity and instruction following even over 10+ interaction rounds.

Related papers

• Thinking on Shots: Consistent Multi-Shot Video Editing with Agentic Reasoning [56.17582599802973]
  We introduce the Multi-Instruction Multi-Shot Long-Video Editing (MMLVE) task, which is structured around three core objectives.<n>To tackle these three unique challenges, we introduce an agentic editing framework that leverages the Large Language Models (LLMs) and Vision-Language Models (VLMs)<n>Extensive experiments demonstrate that our MMLVE-Agent outperforms existing closed-source SOTA approaches.
  arXiv  Detail & Related papers  (2026-08-27T08:48:13Z)
• MSEditor: Toward Consistent Multi-Shot Video Editing [46.964066826513374]
  We propose MSEditor, the first framework designed specifically for consistent multi-shot video editing.<n>To overcome the scarcity of high-quality multi-shot training data, we repurpose existing multi-view video datasets to provide robust cross-shot supervision.<n>To effectively mitigate cumulative errors and ensure long-range temporal coherence, we design a Cross-Shot Packing strategy.
  arXiv  Detail & Related papers  (2026-08-18T09:21:58Z)
• Edit-R2: Context-Aware Reinforcement Learning for Multi-Turn Image Editing [42.176441824728066]
  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.
  arXiv  Detail & Related papers  (2026-06-04T09:49:47Z)
• SAMA: Factorized Semantic Anchoring and Motion Alignment for Instruction-Guided Video Editing [76.349958946335]
  We present SAMA (factorized Semantic Anchoring and Motion Alignment), a framework that factorizes video editing into semantic anchoring and motion modeling.<n>First, we introduce Semantic Anchoring, which establishes a reliable visual anchor by jointly predicting semantic tokens and video latents at sparse anchor frames.<n>Second, Motion Alignment pre-trains the same backbone on motion-centric video restoration pretext tasks.
  arXiv  Detail & Related papers  (2026-03-19T17:59:51Z)
• The Devil is in Attention Sharing: Improving Complex Non-rigid Image Editing Faithfulness via Attention Synergy [71.39358554558667]
  We introduce SynPS, a method that Synergistically leverages Positional embeddings and Semantic information for faithful non-rigid image editing.<n>We propose an editing measurement that quantifies the required editing magnitude at each denoising step.<n>By adaptively integrating positional and semantic cues, SynPS effectively avoids both over- and under-editing.
  arXiv  Detail & Related papers  (2025-12-16T14:08:00Z)
• FreqEdit: Preserving High-Frequency Features for Robust Multi-Turn Image Editing [23.172235115031018]
  We present FreqEdit, a training-free framework that enables stable editing across 10+ consecutive iterations.<n>Our approach comprises three synergistic components: (1) high-frequency feature injection from reference velocity fields to preserve fine-grained details, (2) an adaptive injection strategy that spatially modulates injection strength for precise region-specific control, and (3) a path compensation mechanism that periodically recalibrates the editing trajectory to prevent over-constraint.
  arXiv  Detail & Related papers  (2025-12-01T15:00:47Z)
• Video4Edit: Viewing Image Editing as a Degenerate Temporal Process [24.8621496006791]
  multimodal foundation models have propelled instruction-driven image generation and editing into a genuinely cross-modal, cooperative regime.<n>We revisit this challenge through the lens of temporal modeling.<n>This perspective allows us to transfer single-frame evolution priors from video pre-training, enabling a highly data-efficient fine-tuning regime.
  arXiv  Detail & Related papers  (2025-11-22T17:30:55Z)
• Towards One-step Causal Video Generation via Adversarial Self-Distillation [71.30373662465648]
  Recent hybrid video generation models combine autoregressive temporal dynamics with diffusion-based spatial denoising.<n>Our framework produces a single distilled model that flexibly supports multiple inference-step settings.
  arXiv  Detail & Related papers  (2025-11-03T10:12:47Z)
• Zero-Shot Video Editing through Adaptive Sliding Score Distillation [51.57440923362033]
  This study proposes a novel paradigm of video-based score distillation, facilitating direct manipulation of original video content. We propose an Adaptive Sliding Score Distillation strategy, which incorporates both global and local video guidance to reduce the impact of editing errors.
  arXiv  Detail & Related papers  (2024-06-07T12:33:59Z)

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.