Software & Algorithms
Free and open-source geospatial software from Urban Geo Analytics: edit and animate street scenes with AI, score every street of a city, measure walkable access to population, and turn image collections into data.
RUE: Render Urban Edits
Free desktop application · Windows · runs on your own GPU
RUE (Render Urban Edits) is a free desktop application that turns a photo of a street into a project image or video with AI. Describe the change in plain words or paint the area to change, and RUE gives back the edited image, a short video of the scene coming to life, or a transition from the street as it is today to the project. Everything runs on your own computer, with no account and nothing uploaded.

SAGAI: Streetscape Analysis with Generative AI

Open source · Apache 2.0 · runs in Google Colab
SAGAI is an open-source workflow for scoring and mapping street-level urban environments with vision-language models. From a geographic bounding box, it samples the streets from OpenStreetMap, downloads Google Street View images and scores them with any of 24 models through UVLM, by asking questions in plain language: no labelled data, no deep learning expertise, and nothing to install.
PPCA: Population Potential on Catchment Area
Open protocol · Python · OpenStreetMap + Global Human Settlement data
The PPCA (Population Potential on Catchment Areas) protocol is a four-step procedure for acquiring, processing, and analyzing geospatial data from OpenStreetMap (OSM) and Global Human Settlement (GHS) data using Python and associated libraries. The project aims to evaluate the potential number of people accessible on foot within specified proximity catchment areas. The protocol works globally by requiring only bounding box coordinates instead of sample data.

UVLM: Unified Vision-Language Model Loader

Open source · Apache 2.0 · Python package
UVLM is an open-source Python package that loads and runs 24 vision-language models from five families through one interface. Turn image collections into data and compare models with identical prompts and evaluation protocols, without writing model-specific code.