CLIP the landscape: Automated tagging of crowdsourced landscape images

We present a CLIP-based, multi-modal, multi-label classifier for predicting geographical context tags from landscape photos in the Geograph dataset–a crowdsourced image archive spanning the British Isles, including remote regions lacking POIs and street-level imagery. Our approach addresses a Kaggle competition task based on a subset of Geograph’s 8M images, with strict evaluation: exact match accuracy is required across 49 possible tags. We show that combining location and title embeddings with image features improves accuracy over using image embeddings alone. We release an efficient pipeline 2 2 https://github.com/SpaceTimeLab/ClipTheLandscape . that trains on a modest laptop, using pre-trained CLIP image and text embeddings and a simple classification head. Predicted tags can support downstream tasks such as building location embedders for GeoAI applications, enriching spatial understanding in data-sparse regions. • Tri-modal CLIP model fuses image, title, and location to predict 49 geographic context tags. • Tag prediction difficulty varies widely, reflecting subjectivity and label ambiguity. • Titles outperform image features, revealing annotation bias in crowdsourced data. • Trains on 0.5M images in under 10 min on a laptop; open-source pipeline released. • Scales to all 8M Geograph images, enabling large-scale tagging of crowdsourced images.