Search papers, labs, and topics across Lattice.
This paper addresses the challenge of preserving occluded content in conversational image editing, where previously visible regions may disappear due to modifications. The authors introduce OCCUR-Bench, a benchmark designed to evaluate temporal preservation through diverse scenarios and historical references, and propose ReSpec, a framework that utilizes restoration-aware instructions to enhance fidelity in image restoration. Experimental results demonstrate that ReSpec significantly improves restoration fidelity and temporal consistency, underscoring the importance of grounding preservation in the editing history rather than solely relying on the current image state.
Occluded content in image editing can be accurately restored by grounding preservation in historical context, rather than just the current frame.
Conversational image editing requires preserving not only visible content, but also content that temporarily disappears across turns. When newly added or modified content occludes a previously visible region, that region should reappear if it was never semantically changed. However, existing systems often fail to recover such occluded-but-unchanged content, producing inconsistent or hallucinated results. We introduce OCCUR-Bench, a diagnostic benchmark for temporal preservation in conversational image editing. OCCUR-Bench provides diverse occlusion-and-revelation scenarios with historical restoration references, enabling evaluation of faithful restoration rather than plausible regeneration. We also propose ReSpec, a training-free framework that makes implicit preservation explicit by pairing restoration-aware instructions with historical visual references. Given an editing history, ReSpec identifies what should persist, selects the historical image state that provides missing visual evidence, and conditions an in-context editor on the resulting instruction and reference image. Experiments show that ReSpec improves restoration fidelity and temporal consistency on OCCUR-Bench, highlighting the need to ground preservation in editing history rather than only the current image.
