Skip to content

Acknowledgements and attributions

Valdrada is compiled almost entirely from public records and openly licensed assets. This file lists every external source contained in the repository, the compiled city or the release builds, with the attribution each licence requires. LICENSING.md gives the licence of each part of the project, and client/DATA_SOURCES.md records, field by field, what each dataset contributes.

City data

City of New York, NYC Open Data (NYC Open Data terms of use):

  • Building Footprints (5zhs-2jue)
  • PLUTO (64uk-42ks)
  • DOB NOW: Safety – Facades Compliance Filings (xubg-57si)
  • Street Centerline, CSCL (inkn-q76z)
  • 2015 Street Tree Census (uvpi-gqnh)
  • NYCDEP Citywide Hydrants (5bgh-vtsn)
  • Parks Properties (enfh-gkve)
  • Borough Boundaries (gthc-hcne)
  • Bus Stop Shelters (t4f2-8md7)
  • LinkNYC Kiosks (s4kf-3yrf)
  • DOT Bicycle Parking (592z-n7dk)
  • Bus Lanes – Local Streets (ycrg-ses3)
  • Planimetric Database: Sidewalk (52n9-sdep), Roadbed (i36f-5ih7), Median (ees7-4ufv), Parking Lot (7cgt-uhhz), Public Plazas (ue2e-9jm2) and Open Space, Parks (y6ja-fw4f) and Other (b7j8-z8a7)
  • drinking-water tank inspection records (building identification numbers only)
  • the DoITT NYC 3D Building Model (CityGML LOD2, 2014)

New York State, data.ny.gov.

  • MTA Subway Entrances and Exits, 2024 (i9wp-a4ja)
  • NYSERDA Statewide Distributed Solar Projects (locations, module counts)

OpenStreetMap contributors. The project uses building:colour tags, gap-fill building footprints, the Columbia University campus micro-map and elevated-rail alignments. © OpenStreetMap contributors, available under the Open Database Licence (ODbL 1.0), https://www.openstreetmap.org/copyright. Data derived from OpenStreetMap is contained in the compiled tiles and in client/public/data/, and is available under the ODbL 1.0. The attribution must accompany any distribution of the compiled city or imagery rendered from it.

U.S. Geological Survey. 3DEP elevation (via AWS Terrain Tiles) for park relief. Public domain.

Rooftop inventories.

  • M. L. Treglia et al., green-roof footprints for New York City, The Nature Conservancy (2016).
  • A. Wu and F. Biljecki, Roofpedia: automatic mapping of green and solar roofs for an open roofscape registry and evaluation of urban sustainability, Landscape and Urban Planning 214 (2021), https://github.com/ualsg/Roofpedia.

Federal Highway Administration. Pavement-arrow outlines (MUTCD 2009, Figure 3B-24) and the Standard Alphabets for Traffic Control Devices (Series C) glyph outlines, extracted from the official PDFs. Public domain.

3D models and animation

  • CARLA Simulator (https://carla.org), assets under CC BY 4.0.
  • client/public/models/carla/: vehicles and about 180 street, construction and trash props, converted from CARLA 0.9.15.
  • client/public/models/fleet24/: 16 vehicle models, converted from CARLA 0.10.0.
  • client/public/models/peds24/: 25 pedestrian bodies in 37 outfit variants, converted from CARLA 0.10.0.

NYC-specific edits: taxi and police liveries, New York licence plates, a plain white box-truck body, removed CARLA lettering and recalibrated skin tones. Vehicle and pedestrian models: CARLA Simulator (carla.org), CC BY 4.0.

The pedestrian bank contains components created with Epic Games' MetaHuman. Those components are subject to Epic's MetaHuman licence in addition to CC BY 4.0, which does not allow using them to build or enhance a database or to train or test AI models. See the MetaHuman notice in LICENSING.md and the server option --pedestrians procedural. - Microsoft Rocketbox Avatar Library (https://github.com/microsoft/Microsoft-Rocketbox), MIT licence, Copyright (c) 2020 Microsoft. M. Gonzalez-Franco et al., The Rocketbox Library and the Utility of Freely Available Rigged Avatars, Frontiers in Virtual Reality 1, 2020, doi:10.3389/frvir.2020.561558. - client/public/models/peds24/rb27/: 49 pedestrian bodies (45 adults, 4 children), re-bound to the pedestrian skeleton; the licence text travels with them in rb27/NOTICE.md. - Alma Mater photogrammetry scan by M. K. Turkcan (https://doi.org/10.5281/zenodo.10312053), CC BY 4.0: client/public/models/landmarks/alma_mater.glb, the statue on the Low Library steps (Daniel Chester French, 1903), cut to the bronze, its throne and the stone die and simplified by tools/assets/alma_mater.mjs. - 100STYLE locomotion dataset: I. Mason, S. Starke and T. Komura, Real-Time Style Modelling of Human Locomotion via Feature-Wise Transformations and Local Motion Phases, 2022, Zenodo record 8127870. CC BY 4.0. Sixteen styles (walk and idle) and four further walks are retargeted onto the pedestrian skeleton (peds24/clips_gen2.bin). - Bag and backpack models (Sketchfab, obtained via Objaverse 1.0), each CC BY 4.0: - "Retreat Herschel bag" by alban - "Kanken backpack" by Modelified (recoloured) - "Sling Bag" by Hydro3D Solution - "Feuerwear Shoulder Bag Walter UK" by Feuerwear - "Worn leather handbag" by Lassi Kaukonen

Textures, skies and fonts

  • Poly Haven (https://polyhaven.com), CC0 1.0:
  • PBR sets: asphalt_02, dirty_concrete, leafy_grass, concrete_pavement_03, red_bricks_04, brown_brick_02, white_bricks, sandstone_blocks_04, plastered_wall_05, stone_wall_03, sandstone_blocks_08, concrete_floor_worn_001 (the Columbia campus stone), asphalt_pit_lane, aerial_asphalt_01, gravel_floor_02 (the ground's asphalt and crushed stone)
  • eight 4K sky HDRIs
  • ambientCG (https://ambientcg.com), CC0 1.0: Asphalt025C, Asphalt031, Concrete031, Concrete037, Concrete048, Grass004, PavingStones128, Bricks090.
  • Inter by Rasmus Andersson, SIL Open Font License 1.1 (client/public/fonts/OFL.txt).
  • Google Fonts display families on the Times Square boards and shop signs, twenty families, each under the SIL Open Font License 1.1; client/public/fonts/ta31/FONTS.txt lists them with their copyright holders, and each family's licence text is beside its files.
  • The colour-grade LUTs in client/public/luts/ and the storefront signage are generated in this repository.

Software

Component Licence Use
three.js 0.185 MIT renderer (bundled)
postprocessing 6.39 Zlib post-processing passes (bundled)
n8ao 2.0 ISC ambient occlusion (bundled)
ez-tree 1.1 (Dan Greenheck) MIT parametric trees (bundled)
earcut 3.2 ISC polygon triangulation (bundled, compiler)
meshoptimizer decoder MIT mesh decompression (bundled)
Basis Universal transcoder Apache-2.0 KTX2 texture transcoding (bundled)
Electron 44 (Chromium, Node.js) MIT; Chromium licences in LICENSES.chromium.html simulation server runtime (release builds)
Vite 8, pngjs MIT build and tooling
@electron/packager BSD-2-Clause release packaging
Playwright, puppeteer-core Apache-2.0 headless render harnesses (tooling only)
ffmpeg-static GPL-3.0-or-later video tooling only; not bundled or distributed
Turbo colour map, polynomial fit (Anton Mikhailov, Google) Apache-2.0 depth visualisation in tools/figures/

Design references, both MIT-licensed: BuildingGeneratorThreeJS by achrefelouafi for the precipitation and wet-surface model and the window air-conditioner kit, and Threejs-Awesome-Graphics-Agent-Skills for the precipitation-surface conventions.

Reference imagery

During development, Google Street View panoramas and web photographs of New York buildings were viewed as visual references for comparison renders. They are not contained in this repository, the compiled city or any release. No such image has been used as a texture, as a source of geometry or as training data.

Earlier work and the name

This project was published as BoundlessNYC up to version 0.3.0 and was renamed Valdrada in version 0.3.1, to avoid confusion with the earlier Boundless project, which is a separate work. Valdrada is developed and maintained by Mehmet Kerem Turkcan and acknowledges that earlier work: M. K. Turkcan, Y. Li, C. Zang, J. Ghaderi, G. Zussman and Z. Kostic. Boundless: Generating photorealistic synthetic data for object detection in urban streetscapes. arXiv:2409.03022, 2024. https://arxiv.org/abs/2409.03022

To cite Valdrada itself, use CITATION.cff.