RUE · Render Urban Edits

From a street photo to a project image or video. Free, and it runs on your own computer.

What is RUE?

RUE, short for Render Urban Edits, is a desktop application for anyone who has to show what a street could become: planners, designers, researchers, developers, local authorities. Start from a photo of a street or a public space and describe the change in plain words, or paint the area you want to change. RUE gives you 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.

After a one-time download of the engine and the models, everything runs on your computer. There’s no account, and nothing you create is uploaded: your photos and your projects stay on your machine.

The free tools

Image generation draws a picture from a written description, in seven formats.

Image edit changes a photo with words, and can take up to two reference pictures for a material, an object or a style. Paint a mask and only the painted area is redrawn, such as the roadway, a façade or a parking lot. You choose whether what’s under the mask guides the result or is erased first. The rest of the photo is kept as it was, and the result is cropped to a standard format of about one megapixel.

Video generation makes a two- to five-second clip, in 480p or 720p: people walking, cyclists passing, trees moving in the wind, a slow camera move. Start from a picture, or from a prompt alone, in which case RUE draws the first picture itself.

First to last frame takes the photo of today and the image of the project, and makes the three- to five-second video that goes from one to the other.

Every run goes into a queue, so you can line up several takes, fix a seed to repeat a result, and send any result on to another tool.

Custom modules, on demand

RUE’s core is free. When a project needs more, we build the module when you order it and add it to your own version of RUE.

Scenario presetsTurn a street photo into a planning scenario in one click: greening, pedestrianisation, bike lane, depaving, cool island, tactical urbanism.
Before / afterA comparison slider, side-by-side boards and a short morphing video from today to the project: the change the way residents understand it.
Tree growth & seasonsThe same view at planting, after 10 and after 20 years, in summer, in winter or at night.
Walk-through videoSeveral edited viewpoints chained into one walk along the transformed street.
Materials & façadesStone, brick, wood cladding, green walls and permeable paving from a catalogue.
Street furnitureBenches, lamps, bike racks, bus shelters and planters exactly where you paint them, from a catalogue or your city’s own street-furniture charter.
Clean the photoRemove parked cars, signs, temporary works and text, and blur faces and number plates before publishing (GDPR).
Measure the scenarioScore every before / after with SAGAI and UVLM, for greenery, walkability, shade and enclosure: an argument based on measurements, not only on pictures.
A whole streetApply the scenario to every sampled view along a street, and map the measured change.
Image data extractionTurn thousands of street photos into a research dataset by asking questions in plain words.
Your city’s lookA model trained on your own materials, architecture and street furniture. Your pictures stay on your machines.
Your own toolSomething specific to your project, designed and built into your version of RUE.

Contact us about a module →

Who uses it

Participatory planning teamsBring RUE to public meetings, co-design workshops and online consultations, so residents react to an image rather than a plan.
Urban research labsUse it to build visual stimuli for perception studies, focus groups and stated-preference surveys.
Urban ambiance studiesUse it to vary light, seasons, greenery and shade, and to explore the sense of place.
Real-estate developersUse it for sales presentations, permit applications and marketing videos.
Cities and local authoritiesUse it for council presentations, public consultations and public communication.
Urban planning and design agenciesUse it for feasibility studies, design options and client presentations.

Requirements

You’ll need:

  • Windows, with an NVIDIA GPU with 16 GB of video memory recommended (the setup won’t install below 12 GB);
  • 32 GB of system RAM recommended;
  • about 70 GB of free disk space, of which 50 GB is models; an internet connection for the first launch.

RUE was tested on an ASUS ROG Strix G16 (G614PW) laptop with an NVIDIA GeForce RTX 5080 Laptop GPU (16 GB of video memory) and 32 GB of system RAM. On that machine:

  • the engine takes about 40 seconds to start;
  • an image takes about 25 seconds, and an edit about 40;
  • a four-second video at 480p takes about two minutes.

Installation

Download RUE-Setup-1.0.0.exe and run it. On its first launch, RUE installs its own engine in your user folder: Python, PyTorch, a fixed version of ComfyUI, and the ten model files (about 50 GB). Each download is checked against a fingerprint and picks up where it stopped if it’s interrupted. You can look around the app while it downloads. RUE uses its own engine and leaves any ComfyUI you already have untouched. Recent NVIDIA drivers work best; with an older driver (before 580), RUE installs a compatible build on its own.

Credits

Developed by Urban Geo Analytics (UGA). RUE is built on open-source software and open-weight models, all credited in the app. RUE is free to use, and what you create with it is yours to use as you wish. RUE itself is not open source: its license terms come with the software. Custom modules are offered under a separate agreement, set for each project.

Built by Joan Perez, Kevin Perez and Maurice Stolk. Published by Joan Perez, EI, trading as Urban Geo Analytics.