TR2001-30

Example-based super-resolution


    •  William T. Freeman, Thouis R. Jones, and Egon C. Pasztor, "Example-based super-resolution", Tech. Rep. TR2001-30, Mitsubishi Electric Research Laboratories, Cambridge, MA, August 2001.
      BibTeX TR2001-30 PDF
      • @techreport{MERL_TR2001-30,
      • author = {William T. Freeman, Thouis R. Jones, and Egon C. Pasztor},
      • title = {Example-based super-resolution},
      • institution = {MERL - Mitsubishi Electric Research Laboratories},
      • address = {Cambridge, MA 02139},
      • number = {TR2001-30},
      • month = aug,
      • year = 2001,
      • url = {https://www.merl.com/publications/TR2001-30/}
      • }
  • Research Areas:

    Artificial Intelligence, Computer Vision, Machine Learning

Abstract:

Image-based models for computer graphics lack resolution independence: they cannot be zoomed much beyond the pixel resolution they were sampled at without a degradation of quality. Interpolating images usually results in a blurring of edges and image details. We describe image interpolation algorithms which use a database of training images to create plausible high-frequency details in zoomed images. Image pre-processing steps allow the use of image detail from regions of the training images which may look quite different from the image to be processed. These methods preserve fine details, such as edges, generate believable textures, and can give good results even after zooming multiple octaves.

 

  • Related News & Events

    •  AWARD    MERL paper wins major award from IEEE Computer Society
      Date: January 12, 2023
      Awarded to: William T. Freeman, Thouis R. Jones, and Egon C. Pasztor
      Awarded by: IEEE Computer Society
      Research Areas: Computer Vision, Machine Learning
      Brief
      • The MERL paper entitled, "Example-Based Super-Resolution" by William T. Freeman, Thouis R. Jones, and Egon C. Pasztor, published in a 2002 issue of IEEE Computer Graphics and Applications, has been awarded a 2021 Test of Time Award by the IEEE Computer Society. This work was done while the principal investigator, Prof. Freeman, was a research scientist at MERL; he is now a Professor of Electrical Engineering and Computer Science at MIT.

        This best paper award recognizes regular or special issue papers published by the magazine that have made profound and long-lasting research impacts in bridging the theory and practice of computer graphics. "This paper is an early example of using learning for a low-level vision task and we are very proud of the pioneering work that MERL has done in this area prior to the deep learning revolution," says Anthony Vetro, VP & Director at MERL.
    •