[论文] HoloGeo: Mitigating Landmark Bias in Geo-localization via Evidence-Driven Reasoning

论文概要

研究领域: cs.CV 作者: Pengcheng Zhou, Xuanyu Liu, Yanchen Yin, Bobo Li, Shengqiong Wu, Mong-Li Lee, Wynne Hsu 发布时间: 2026-07-16 arXiv: 2607.15255

中文摘要

视觉语言模型（VLM）在图像地理定位任务上取得了显著进展，但现有模型容易受到地标偏差的干扰——它们会忽略真正的地理线索，或形成虚假的地标关联，导致定位结果不准确。为系统研究这一问题，作者首先提出了两个量化指标：偏差强度（BI）和偏差危害性（BH），用于刻画地标对模型推理的影响，并建立了综合基准 LandmarkBias-3K。为缓解地标偏差，他们进一步提出了证据驱动推理框架 HoloGeo，通过高质量数据集 BF-30k（标注了结构化、无偏差的多证据推理链）进行支撑。HoloGeo 引入多维度奖励机制，显式鼓励模型均衡关注多样化的视觉线索，实现证据驱动的联合推理。大量实验表明，HoloGeo 在 IM2GPS3K 和 YFCC4k 上保持优秀性能的同时，在 LandmarkBias-3K 上显著优于现有开源 VLM，验证了其在鲁棒地理空间推理方面的有效性。

原文摘要

Recent advances in Vision-Language Models (VLMs) have significantly improved image geo-localization, yet existing models remain susceptible to landmark bias, causing them to overlook geographical cues or form spurious correlations, ultimately resulting in inaccurate localization. To systematically investigate this issue, we first design two quantitative metrics, Bias Intensity (BI) and Bias Harmfulness (BH), to characterize the impact of landmarks exerted on model reasoning, and establish a comprehensive benchmark, LandmarkBias-3K. To mitigate landmark bias, we further propose an evidence-driven reasoning framework, HoloGeo, to improve the reliability of geo-localization. HoloGeo is supported by a high-quality dataset, BF-30k, annotated with structured multi-evidence bias-free reasoning chains. By incorporating multi-dimensional rewards, HoloGeo explicitly encourages balanced attention over diverse visual cues and achieves evidence-driven joint reasoning. Extensive experiments demonstrate that HoloGeo not only maintains excellent performance on IM2GPS3K and YFCC4k but also significantly outperforms existing open-source VLMs on LandmarkBias-3K, validating its effectiveness for robust geospatial reasoning.

--- *自动采集于 2026-07-20*

#论文 #arXiv #AI #小凯