J. Mairal, F. Bach, J. Ponce, G. Sapiro and A. Zisserman. Discriminative Learned Dictionaries for Local Image Analysis. IEEE Conference on Computer Vision and Pattern Recognition, 2008.
J. Mairal, M. Leordeanu, F. Bach, M. Hebert and J. Ponce. Discriminative Sparse Image Models for Class-Specific Edge Detection and Image Interpretation. European Conference on Computer Vision, 2008.
Both papers go together and they describe one way to learn and use sparse coding techniques. Although the techniques were originally intended for image processing, e.g., compression and coding, there are some ideas on how to use them for recognition.
Subscribe to:
Post Comments (Atom)
Vishnu, can you post the links that were in your presentation, near the very end?
ReplyDeleteI'm especially interested in some of the in-painting stuff you mentioned. Can a similar approach be used for boosting the resolution of an image? I don't remember off the top of my head if this is how the "hallucination" method worked.
Vishu, again, very nice presentation.
ReplyDeleteI'm interested in having the PDF of your presentation, if it is possible.
For those who want to know more about this topic, i discovered that a good name to search, in the area of the greedy algorithms, is Joel Tropp.
http://www.acm.caltech.edu/~jtropp/
Martial is going to put up the presentation on the website, the ppt has a slide with the references.
ReplyDelete@Leland, You are absolutely right that we can use this for super-resolution, in fact that is exactly the project I am doing for the course.
@Leland, The in-painting example is described in the following paper
ReplyDeletehttp://www.cs.technion.ac.il/~elad/publications/journals/2006/KSVD_Color_IEEE_TIP.pdf