mlsanityData to 7 Oct 2026

LightCity: An Urban Dataset for Outdoor Inverse Rendering and Reconstruction under Multi-illumination Conditions

arXiv 2602.01118 ↗First posted 1 Feb 2026cs.CVcited by 4

Not on the map. Why?

Abstract (arXiv)

Inverse rendering in urban scenes is pivotal for applications like autonomous driving and digital twins. Yet, it faces significant challenges due to complex illumination conditions, including multi-illumination and indirect light and shadow effects. However, the effects of these challenges on intrinsic decomposition and 3D reconstruction have not been explored due to the lack of appropriate datasets. In this paper, we present LightCity, a novel high-quality synthetic urban dataset featuring diverse illumination conditions with realistic indirect light and shadow effects. LightCity encompasses over 300 sky maps with highly controllable illumination, varying scales with street-level and aerial perspectives over 50K images, and rich properties such as depth, normal, material components, light and indirect light, etc. Besides, we leverage LightCity to benchmark three fundamental tasks in the urban environments and conduct a comprehensive analysis of these benchmarks, laying a robust foundation for advancing related research.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 58 of its 75 references that are in the map. First posted on arXiv 1 Feb 2026.

Builds on

58 of 75 references.

References · 58

  1. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  2. 20233D Gaussian Splatting for Real-Time Radiance Field Renderingcited by 6,141
  3. 2022InstructPix2Pix: Learning to Follow Image Editing Instructionscited by 2,591
  4. 2019Scalability in Perception for Autonomous Driving: Waymo Open Datasetcited by 2,436
  5. 2020NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collectionscited by 1,230
  6. 2023K-Planes: Explicit Radiance Fields in Space, Time, and Appearancecited by 742
  7. 2020Hypersim: A Photorealistic Synthetic Dataset for Holistic Indoor Scene Understandingcited by 593
  8. 2021KITTI-360: A Novel Dataset and Benchmarks for Urban Scene Understanding in 2D and 3Dcited by 570
  9. 2020NeRV: Neural Reflectance and Visibility Fields for Relighting and View Synthesiscited by 493
  10. 2020NeRD: Neural Reflectance Decomposition from Image Collectionscited by 446
  11. 2021Mega-NeRF: Scalable Construction of Large-Scale NeRFs for Virtual Fly-Throughscited by 396
  12. 2021PhySG: Inverse Rendering with Spherical Gaussians for Physics-based Material Editing and Relightingcited by 332
38 more
  1. 2023OmniObject3D: Large-Vocabulary 3D Object Dataset for Realistic Perception, Reconstruction and Generationcited by 324
  2. 2023GS-IR: 3D Gaussian Splatting for Inverse Renderingcited by 220
  3. 2024CityGaussian: Real-time High-quality Large-Scale Scene Rendering with Gaussianscited by 190
  4. 2023MatrixCity: A Large-scale City Dataset for City-scale Neural Rendering and Beyondcited by 185
  5. 2022Modeling Indirect Illumination for Inverse Renderingcited by 167
  6. 2023TensoIR: Tensorial Inverse Renderingcited by 166
  7. 2018Deep Appearance Models for Face Renderingcited by 164
  8. 2024RGBleftrightarrowX: Image decomposition and synthesis using material- and lighting-aware diffusion modelscited by 159
  9. 2024StableNormal: Reducing Diffusion Variance for Stable and Sharp Normalcited by 153
  10. 2024WildGaussians: 3D Gaussian Splatting in the Wildcited by 150
  11. 2023Relightable 3D Gaussians: Realistic Point Cloud Relighting with BRDF Decomposition and Ray Tracingcited by 138
  12. 2021Hallucinated Neural Radiance Fields in the Wildcited by 130
  13. 2021Capturing, Reconstructing, and Simulating: the UrbanScene3D Datasetcited by 117
  14. 2024Gaussian in the Wild: 3D Gaussian Splatting for Unconstrained Image Collectionscited by 117
  15. 2021NeRF for Outdoor Scene Relightingcited by 114
  16. 2024One-Step Image Translation with Text-to-Image Modelscited by 111
  17. 2023Neural Fields meet Explicit Geometric Representation for Inverse Rendering of Urban Scenescited by 89
  18. 2018CGIntrinsics: Better Intrinsic Image Decomposition through Physically-Based Renderingcited by 85
  19. 2022Neural 3D Reconstruction in the Wildcited by 79
  20. 2018Learning Intrinsic Image Decomposition from Watching the Worldcited by 74
  21. 2023Intrinsic Image Diffusion for Indoor Single-view Material Estimationcited by 73
  22. 2022Learning to Relight Portrait Images via a Virtual Light Stage and Synthetic-to-Real Adaptationcited by 67
  23. 2024IntrinsicAnything: Learning Diffusion Priors for Inverse Rendering Under Unknown Illuminationcited by 67
  24. 2024SWAG: Splatting in the Wild images with Appearance-conditioned Gaussianscited by 61
  25. 2024Wild-GS: Real-Time Novel View Synthesis from Unconstrained Photo Collectionscited by 56
  26. 2023Cross-Ray Neural Radiance Fields for Novel-view Synthesis from Unconstrained Image Collectionscited by 54
  27. 2023Intrinsic Image Decomposition via Ordinal Shadingcited by 53
  28. 2023Stanford-ORB: A Real-World 3D Object Inverse Rendering Benchmarkcited by 51
  29. 2022IntrinsicNeRF: Learning Intrinsic Neural Radiance Fields for Editable Novel View Synthesiscited by 50
  30. 2021Self-supervised Outdoor Scene Relightingcited by 48
  31. 2019A Dataset of Multi-Illumination Images in the Wildcited by 45
  32. 2024Colorful Diffuse Intrinsic Image Decomposition in the Wildcited by 44
  33. 2016Learning Non-Lambertian Object Intrinsics across ShapeNet Categoriescited by 42
  34. 2023ReLight My NeRF: A Dataset for Novel View Synthesis and Relighting of Real World Objectscited by 42
  35. 2024SplatLoc: 3D Gaussian Splatting-based Visual Localization for Augmented Realitycited by 42
  36. 2023Exploiting Diffusion Prior for Generalizable Dense Predictioncited by 37
  37. 2023A Large-Scale Outdoor Multi-modal Dataset and Benchmark for Novel View Synthesis and Implicit Scene Reconstructioncited by 32
  38. 2023OpenIllumination: A Multi-Illumination Dataset for Inverse Rendering Evaluation on Real Objectscited by 31

50 most cited of 58 All 58 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 4.

4 papers.

  1. 2026Generative World Renderercited by 8
  2. 2026A 3D Reconstruction Benchmark for Asset Inspectioncited by 0
  3. 2026Wild3R: Feed-Forward 3D Gaussian Splatting from Unconstrained Sparse Photo Collectioncited by 0
  4. 2026BRDFusion: Physics Meets Generation for Urban Scene Inverse Renderingcited by 0