HoloGeo: Mitigating Landmark Bias in Geo-localization via Evidence-Driven Reasoning
2026-07-16 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors studied why current vision-language models for image geo-localization often get confused by famous landmarks, which leads to mistakes. They created two new ways to measure how much these landmarks bias the models and built a special benchmark called LandmarkBias-3K to test this problem. To fix the bias, the authors introduced HoloGeo, a system trained on a new dataset with clear, unbiased explanations. HoloGeo looks at many different clues in images before deciding where a photo was taken, making it more accurate and reliable. Experiments showed that HoloGeo works better than existing models, especially in avoiding landmark bias.
Vision-Language ModelsImage Geo-localizationLandmark BiasBias IntensityBias HarmfulnessBenchmark DatasetEvidence-driven ReasoningHoloGeoMulti-evidence ReasoningGeospatial Reasoning
Authors
Pengcheng Zhou, Xuanyu Liu, Yanchen Yin, Bobo Li, Shengqiong Wu, Mong-Li Lee, Wynne Hsu
Abstract
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.