UrbanGraphEmbeddings: Learning and Evaluating Spatially Grounded Vision–Language Embeddings for Urban Environments DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY ISBN: 978-1-4503-XXXX-X/2018/06 Tha

Jie Zhang ∗, Xingtong Yu ∗, Yuan Fang †, Rudi Stouffs †, Zdravko Trivic † · 2026

Read UrbanGraphEmbeddings: Learning and Evaluating Spatially Grounded Vision–Language Embeddings for Urban Environments DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY ISBN: 978-1-4503-XXXX-X/2018/06 Tha by Jie Zhang ∗, Xingtong Yu ∗, Yuan Fang †, Rudi Stouffs †, Zdravko Trivic † (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-08342.txt.

Description

Street photos alone miss the structure of a city: which streets connect, how far apart places are, and what sits nearby. Jie Zhang and four co-authors build UGData, a dataset that ties street-view images to spatial graphs, plus a training method and a benchmark for city understanding. Their 2026 arXiv paper argues that urban AI needs explicit geography, not just pixels.

The training method, called UGE, aligns images, text, and graph structure in two stages, using instruction-guided contrast and graph encoding. The benchmark tests geolocation, retrieval, perception, and grounding. Built on a seven-billion-parameter vision-language backbone, UGE lifts image retrieval by up to 44 percent and geolocation ranking by 30 percent on training cities, with strong gains on unseen cities too. Readers get a full stack for spatially aware urban models.

Catalog record

CREATORJie Zhang ∗, Xingtong Yu ∗, Yuan Fang †, Rudi Stouffs †, Zdravko Trivic †
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE74 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-08342

Browse the full library · About the Tower of Babel