Knowledge-Intensive Visual Grounding (KVG) requires models to localize objects using fine-grained, domain-specific entity names rather than generic referring expressions. Although Multimodal Large Language Models (MLLMs) possess rich entity knowledge and strong generic grounding capabilities, they often fail to effectively utilize such knowledge when grounding specialized concepts, revealing a knowledge-grounding gap between internal knowledge and grounding predictions. To address this challenge, we propose a knowledge-aware training paradigm for KVG. Our approach first constructs knowledge-guided reasoning data to encourage models to activate domain-relevant entity knowledge during grounding, and then introduces KARL, a Knowledge-Aware Reinforcement Learning framework that adaptively modulates reward signals according to the model's estimated knowledge mastery of different entities. To facilitate systematic evaluation, we introduce KVG-Bench, a benchmark spanning 10 domains with 1.3K curated test cases covering 531 images and 882 entities. Extensive experiments show that our approach consistently outperforms a wide range of baseline models and achieves substantially stronger cross-domain generalization on unseen categories. The data, codes, and models are released at https://github.com/thunlp/KARL.
@article{arxiv.2503.12797,
title = {KARL: Knowledge-Aware Reasoning and Reinforcement Learning for Knowledge-Intensive Visual Grounding},
author = {Xinyu Ma and Ziyang Ding and Zhicong Luo and Chi Chen and Zonghao Guo and Derek F. Wong and Zhen Zhao and Xiaoyi Feng and Maosong Sun},
journal= {arXiv preprint arXiv:2503.12797},
year = {2026}
}