De-Emphasis of Distracting Image Regions
Using Texture Power Maps

Sara L. Su Frédo Durand Maneesh Agrawala
MIT CSAIL MIT CSAIL Microsoft Research
In Proceedings of Texture 2005, at ICCV 2005

Figure 1: We compute the texture power map (right) of the input image (left). The false-color power map shows the high-frequency distribution of the image, that is, the texture boundaries. We use these higher-order features to reduce the spatial variation of texture in the image.

ABSTRACT

We present a post-processing technique that selectively reduces the salience of distracting regions in an image. Computational models of attention predict that texture variation influences bottom-up attention mechanisms. Our method reduces the spatial variation of texture using power maps, high-order features describing local frequency content in an image. Modification of power maps results in effective regional de-emphasis. We validate our results quantitatively via a human subject search experiment and qualitatively with eye tracking data.

FILES

Sara L. Su, Frédo Durand, and Maneesh Agrawala. De-Emphasis of Distracting Image Regions Using Texture Power Maps. In Texture 2005: Proceedings of the 4th ICCV Workshop on Texture Analysis and Synthesis, pp. 119-124, Beijing, China, October 2005

Paper: PDF

@inproceedings{Su:05:Texture,
  author = "Sara L. Su and Fr\'edo Durand and Maneesh Agrawala",
  title = "De-Emphasis of Distracting Image Regions Using Texture Power Maps",
  booktitle = "Texture 2005: Proceedings of the 4th IEEE International Workshop
              on Texture Analysis and Synthesis in conjunction with ICCV'05",
  pages = "119--124",
  location = "Beijing, China",
  month = "October",
  year = "2005",
}

ACKNOWLEDGEMENTS

We thank the MIT Graphics Group and anonymous reviewers for feedback; Paul Green and Eric Chan for invaluable assistance with data acquisition and analysis; and Aude Oliva, Ruth Rosenholtz, and their students for use of their eyetracker. This work was supported by NSF under Grant No. 0429739 and the Graduate Research Fellowship Program, MIT Project Oxygen, and the Royal Dutch/Shell Group.