Learning low-level vision

    •  William T. Freeman, Egon C. Pasztor, "Learning low-level vision", Tech. Rep. TR99-12, Mitsubishi Electric Research Laboratories, Cambridge, MA, July 1999.
      BibTeX TR99-12 PDF
      • @techreport{MERL_TR99-12,
      • author = {William T. Freeman, Egon C. Pasztor},
      • title = {Learning low-level vision},
      • institution = {MERL - Mitsubishi Electric Research Laboratories},
      • address = {Cambridge, MA 02139},
      • number = {TR99-12},
      • month = jul,
      • year = 1999,
      • url = {}
      • }
  • Research Areas:

    Artificial Intelligence, Computer Vision, Machine Learning


We show a learning-based method for low-level vision problems--estimating scenes from images. We generate a synthetic world of scenes and their corresponding rendered images. We model that world with a Markov network, learning the network parameters from the examples. Bayesian belief propagation allows us to efficiently find a local maximum of the posterior probability for the scene, given the image. We call this approach VISTA--Vision by Image/Scene TrAining. We apply VISTA to the \"super-resolution\" problem (estimating high frequency details from a low-resolution image), showing good results. For the motion estimation problem, we show figure/ground discrimination, solution of the aperture problem, and filling-in arising from application of the same probabilistic machinery.