Reasoning-Guided Part-Level Visual Grounding via Reinforcement Learning
Improves part-level visual grounding for multimodal LLMs, a known bottleneck in fine-grained understanding.
Multimodal LLMs fail at part-level grounding due to missing object-part hierarchy. The proposed OP-HRG method, trained with part-aware GRPO, achieves 4B model outperforming 7B grounding LLMs and SAM3 on PascalPart, PartImageNet, and InstructPart.
Multimodal large language models (MLLMs) ground whole objects well from free-form language queries, but they struggle when the query names a part rather than the object. We trace this to a missing object-part hierarchy, since parts are localized in the same single step used for objects. We propose Object-Part Hierarchical Reflective Grounding (OP-HRG), a coarse-to-fine reasoning-guided grounding strategy that first localizes the parent object and then the part within it. A self-check then reflects on the result, with an extension to re-encode the predicted crop to inspect the region it is correcting. We introduce a part-aware GRPO framework to train our pipeline with stage-wise rewards. A 4B model trained this way outperforms 7B grounding LLMs and SAM3 across PascalPart, PartImageNet, and InstructPart, and transfers to reasoning segmentation.