Perception in Reflection
This addresses limitations in multimodal AI for tasks requiring complex reasoning, though it appears incremental as an enhancement to existing LVLM frameworks.
The paper tackles the problem of imperfect visual perception in large vision-language models by proposing a dual-model reflection mechanism called RePer that enables iterative refinement of perception, achieving quantifiable improvements in image understanding, captioning precision, and hallucination reduction.
We present a perception in reflection paradigm designed to transcend the limitations of current large vision-language models (LVLMs), which are expected yet often fail to achieve perfect perception initially. Specifically, we propose Reflective Perception (RePer), a dual-model reflection mechanism that systematically alternates between policy and critic models, enables iterative refinement of visual perception. This framework is powered by Reflective Perceptual Learning (RPL), which reinforces intrinsic reflective capabilities through a methodically constructed visual reflection dataset and reflective unlikelihood training. Comprehensive experimental evaluation demonstrates RePer's quantifiable improvements in image understanding, captioning precision, and hallucination reduction. Notably, RePer achieves strong alignment between model attention patterns and human visual focus, while RPL optimizes fine-grained and free-form preference alignment. These advancements establish perception in reflection as a robust paradigm for future multimodal agents, particularly in tasks requiring complex reasoning and multi-step manipulation.