Welcome to Urban Geo Analytics
Urban Geo Analytics brings together geography, machine learning, and AI to help cities make sense of the data that shapes them. Founded by Joan Perez, a geographer and researcher specializing in spatial analysis and geospatial AI, UGA develops open-source tools — from vision-language pipelines for streetscape analysis to population and accessibility models — used in Horizon Europe research consortia and international consulting assignments, including for the World Bank. Our mission: give planners, researchers, and policymakers scalable, transparent, data-driven ways to understand and improve the urban environment.
Welcome to Urban Geo Analytics
Urban Geo Analytics brings together geography, machine learning, and AI to help cities make sense of the data that shapes them. Founded by Joan Perez, a geographer and researcher specializing in spatial analysis and geospatial AI, UGA develops open-source tools — from vision-language pipelines for streetscape analysis to population and accessibility models — used in Horizon Europe research consortia and international consulting assignments, including for the World Bank. Our mission: give planners, researchers, and policymakers scalable, transparent, data-driven ways to understand and improve the urban environment.
Our Expertise & Services
Training & Knowledge Sharing
Urban Geo Analytics also offers tailored training and consulting services to empower professionals and organizations with the skills to lead in geospatial innovation:
Check the Last Posts of our Scientific Blog
UVLM v3.2.0 — InternVL3.5 Joins the Registry, With Zero Notebook Changes
UVLM v3.2.0 adds InternVL3.5 (1B–38B, six checkpoints): 21 open VLM checkpoints across 4 families, one Python interface. The new family appeared in the notebooks without a single notebook edit — plus per-model output files for cleaner benchmarking.
The AI Reading Series · From Perceptrons to Agents · Lecture 2: From Neurons to Machines That Talk: How Connectionism Conquered Language
Part 2 of the AI reading series. The previous post ended with AlexNet's 2012 earthquake in image recognition. This one tells the road to language: how researchers turned words into vectors, taught networks to read sequences, discovered attention — and why, in 2017, eight Google researchers decided attention was all you need.
UVLM v3.1.0 — Qwen3-VL Joins the Registry, With Family-Based Model Selection
UVLM v3.1.0 adds a third model family, Qwen3-VL (2B–32B Instruct), bringing the registry to 15 checkpoints. The notebooks gain a two-level family/model selector, the loader picks BF16 automatically on capable GPUs, and the smallest new model runs in about 2 GB of VRAM. Same three-block workflow, same prompts, one more family to compare.