EVLM: Self-Reflective Multimodal Reasoning for Cross-Dimensional Visual Editing
This addresses the challenge of inconsistent edits in vision-language modeling for users needing precise visual content manipulation, representing a novel method for a known bottleneck.
The paper tackles the problem of editing visual content from ambiguous instructions by introducing EVLM, a system that uses reflective reasoning to interpret intent and produce precise editing prompts, achieving substantial gains in alignment with human intent across various editing tasks.
Editing complex visual content from ambiguous or partially specified instructions remains a core challenge in vision-language modeling. Existing models can contextualize content but often fail to infer the underlying intent within a reference image or scene, leading to inconsistent or misaligned edits. We introduce the Editing Vision-Language Model (EVLM), a system that interprets ambiguous instructions in conjunction with reference visuals to produce precise, context-aware editing prompts. EVLM's key innovation is a reflective reasoning framework that translates subjective user intent into structured, actionable outputs by aligning with human-rated rationales through Reflection-Aware KL-Divergence Target Optimization (RKTO). By combining Chain-of-Thought (CoT) reasoning with RKTO alignment, EVLM captures fine-grained editing preferences without relying on binary supervision. Trained on a dataset of 30,000 CoT examples with human-annotated rationale quality, EVLM achieves substantial gains in alignment with human intent. Experiments across image, video, 3D, and 4D editing tasks show that EVLM generates coherent and high-quality instructions, providing a scalable foundation for multimodal editing and reasoning.