[论文] HoloGeo: Mitigating Landmark Bias in Geo-localization via Evidenc...

论文概要

研究领域: CV 作者: Pengcheng Zhou, Xuanyu Liu, Yanchen Yin 发布时间: 2025-07-16 arXiv: 2507.12513

中文摘要

视觉语言模型（VLM）的最新进展显著提升了图像地理定位能力，但现有模型仍易受地标偏见影响——它们会忽视地理线索或形成虚假关联，最终导致定位不准确。为系统研究这一问题，本文首先设计了两个定量指标：偏见强度（BI）和偏见危害性（BH），以刻画地标对模型推理的影响，并建立了综合性基准数据集 LandmarkBias-3K。为缓解地标偏见，进一步提出基于证据驱动的推理框架 HoloGeo，以提升地理定位的可靠性。HoloGeo 依托高质量数据集 BF-30k，该数据集标注了结构化的多证据无偏见推理链。通过引入多维奖励机制，HoloGeo 显式鼓励对多样化视觉线索的均衡关注，实现证据驱动的联合推理。大量实验表明，HoloGeo 不仅在 IM2GPS3K 和 YFCC4k 上保持优异性能，还在 LandmarkBias-3K 上显著优于现有开源 VLM，验证了其在稳健地理空间推理中的有效性。

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 o...

--- *自动采集于 2026-07-19*

#论文 #arXiv #CV #小凯