Xavier Bresson

 

Last resume (July 2011)

 

 


My current research

 

Convex Relaxation Methods (CRM) for Image Processing and Machine Learning: convex optimization techniques to find (exact or tight approximations of) solutions of non-convex geometric problems in imaging and machine learning.

For an overview of the CRM, see my tutorial with Thomas Pock at the 2011 IEEE International Conference on Image Processing link.

Energy Unification in Image Processing: unifying image processing problems into a single energy model via the Polyakov model.



 

My previous webpage at UCLA (including codes)

 



Prospective Postdocs
I am looking for a Postdoc in the field of Image Processing or Machine Learning.
If you are interested, please contact me by email and include your CV.




Recent papers:

 

- Multi-class unsupervised clustering

Xavier Bresson and Thomas Laurent, "Asymmetric Cheeger cut and application to multi-class unsupervised clustering", pdf

 

- CRM applied to the transductive learning problem

Xavier Bresson, Xue-Cheng Tai, Tony F. Chan, Arthur Szlam, "Multi-Class Transductive Learning based on $\ell^1$ Relaxations of Cheeger Cut and Mumford-Shah-Potts Model", pdf

Code: zip1, zip2, zip3