mlsanityData to 7 Oct 2026

Invariant Descriptors for Intrinsic Reflectance Optimization

arXiv 2204.04076 ↗First posted 8 Apr 2022cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Intrinsic image decomposition aims to factorize an image into albedo (reflectance) and shading (illumination) sub-components. Being ill-posed and under-constrained, it is a very challenging computer vision problem. There are infinite pairs of reflectance and shading images that can reconstruct the same input. To address the problem, Intrinsic Images in the Wild provides an optimization framework based on a dense conditional random field (CRF) formulation that considers long-range material relations. We improve upon their model by introducing illumination invariant image descriptors: color ratios. The color ratios and the reflectance intrinsic are both invariant to illumination and thus are highly correlated. Through detailed experiments, we provide ways to inject the color ratios into the dense CRF optimization. Our approach is physics-based, learning-free and leads to more accurate and robust reflectance decompositions.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 12 of its 53 references that are in the map. First posted on arXiv 8 Apr 2022.

Builds on

12 of 53 references.

References · 12

  1. 2017Stacked Conditional Generative Adversarial Networks for Jointly Learning Shadow Detection and Shadow Removalcited by 160
  2. 2018CGIntrinsics: Better Intrinsic Image Decomposition through Physically-Based Renderingcited by 85
  3. 2018Learning Intrinsic Image Decomposition from Watching the Worldcited by 74
  4. 2019Unsupervised Learning for Intrinsic Image Decomposition from a Single Imagecited by 55
  5. 2016Learning Non-Lambertian Object Intrinsics across ShapeNet Categoriescited by 42
  6. 2019STAR: A Structure and Texture Aware Retinex Modelcited by 41
  7. 2015Direct Intrinsics: Learning Albedo-Shading Decomposition by Convolutional Regressioncited by 32
  8. 2018Joint Learning of Intrinsic Images and Semantic Segmentationcited by 26
  9. 2017CNN based Learning using Reflection and Retinex Models for Intrinsic Image Decompositioncited by 19
  10. 2016Reflectance Adaptive Filtering Improves Intrinsic Image Estimationcited by 18
  11. 2019ShadingNet: Image Intrinsics by Fine-Grained Shading Decompositioncited by 12
  12. 2020Deep intrinsic decomposition trained on surreal scenes yet with realistic light effectscited by 3

Built on by

Not cited yet.