Bayesian Nonparametric Intrinsic Image Decomposition

  • Jason Chang
  • Randi Cabezas
  • John W. FisherIII
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8692)


We present a generative, probabilistic model that decomposes an image into reflectance and shading components. The proposed approach uses a Dirichlet process Gaussian mixture model where the mean parameters evolve jointly according to a Gaussian process. In contrast to prior methods, we eliminate the Retinex term and adopt more general smoothness assumptions for the shading image. Markov chain Monte Carlo sampling techniques are used for inference, yielding state-of-the-art results on the MIT Intrinsic Image Dataset.


Intrinsic images Dirichlet process Gaussian process MCMC 


  1. 1.
    Barron, J., Malik, J.: Shape, illumination, and reflectance from shading. Tech. rep. Univeristy of California, Berkeley (2013)Google Scholar
  2. 2.
    Barrow, H., Tenenbaum, J.: Recovering intrinsic scene characteristics from images. Computer Vision Systems (1978)Google Scholar
  3. 3.
    Blake, A.: Boundary conditions for lightness computation in Mondrian world. Computer Vision, Graphics, and Image Processing (1985)Google Scholar
  4. 4.
    Chang, J.: Sampling in Computer Vision and Bayesian Nonparametric Mixtures. Ph.D. thesis, Massachusetts Institute of Technology (2014)Google Scholar
  5. 5.
    Chang, J., Fisher III, J.W.: Parallel sampling of DP mixture models using sub-clusters splits. In: Neural Information and Processing Systems (December 2013)Google Scholar
  6. 6.
    Freeman, W.T., Pasztor, E.C., Carmichael, O.T.: Learning low-level vision (2000)Google Scholar
  7. 7.
    Funt, B.V., Drew, M.S., Brockington, M.: Recovering shading from color images. In: Sandini, G. (ed.) ECCV 1992. LNCS, vol. 588, pp. 124–132. Springer, Heidelberg (1992)Google Scholar
  8. 8.
    Gehler, P.V., Carsten, R., Kiefel, M., Zhang, L., Schölkopf, B.: Recovering intrinsic images with a global sparsity prior on reflectance. In: Advances in Neural Information Processing Systems (2011)Google Scholar
  9. 9.
    Grosse, R., Johnson, M.K., Adelson, E., Freeman, W.T.: A ground-truth dataset and baseline evaluations for intrinsic image algorithms. In: International Conference on Computer Vision (2009)Google Scholar
  10. 10.
    Hastings, W.K.: Monte Carlo sampling methods using Markov chains and their applications. Biometrika 57(1), 97–109 (1970)CrossRefzbMATHGoogle Scholar
  11. 11.
    Horn, B.: Robot Vision. MIT Press, Cambridge (1986)Google Scholar
  12. 12.
    Land, E., McCann, J.: Lightness and retinex theory. Journal of the Optical Society of America (1971)Google Scholar
  13. 13.
    Malioutov, D., Johnson, J., Choi, M., Willsky, A.: Low-rank variance approximation in gmrf models: Single and multiscale approaches. IEEE Transacations on Signal Processing 56(10), 4621–4634 (2008)CrossRefMathSciNetGoogle Scholar
  14. 14.
    Matsushita, Y., Nishino, I., Ikeuchi, K., Sakauchi, M.: Illumination normalization with time-dependent intrinsic images for video surveillance. IEEE Transacations on Pattern Analysis and Machine Intelligence 26(10), 1336–1347 (2004)CrossRefGoogle Scholar
  15. 15.
    Rasmussen, C.E., Williams, C.K.I.: Gaussian Processes for Machine Learning. MIT Press, Cambridge (2006)zbMATHGoogle Scholar
  16. 16.
    Sethuraman, J.: A constructive definition of Dirichlet priors. Statstica Sinica, 639–650 (1994)Google Scholar
  17. 17.
    Shen, L., Tan, P., Lin, S.: Intrinsic image decomposition with non-local texture cues. In: Computer Vision and Pattern Recognition (2008)Google Scholar
  18. 18.
    Shen, L., Yeo, C.: Intrinsic images decomposition using a local and global sparse representation of reflectance. In: Computer Vision and Pattern Recognition (2011)Google Scholar
  19. 19.
    Sollich, P., Williams, C.K.I.: Using the equivalent kernel to understand Gaussian process regression. In: Advances in Neural Information Processing Systems (2005)Google Scholar
  20. 20.
    Tappen, M.F., Freeman, W.T., Adelson, E.H.: Recovering intrinsic images from a single image. IEEE Transacations on Pattern Analysis and Machine Intelligence 27(9), 1459–1472 (2005)CrossRefGoogle Scholar
  21. 21.
    Weiss, Y.: Deriving intrinsic images from image sequences. In: International Conference on Computer Vision (2001)Google Scholar
  22. 22.
    Zhao, Q., Tan, P., Dai, Q., Shen, L., Wu, E., Lin, S.: A closed-form solution to retinex with nonlocal texture constraints. IEEE Transactions on Pattern Analysis and Machine Intelligence (2012)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2014

Authors and Affiliations

  • Jason Chang
    • 1
  • Randi Cabezas
    • 1
  • John W. FisherIII
    • 1

Personalised recommendations