Ulugbek S. Kamilov

PhD candidate

Biomedical Imaging Group
École polytechnique fédérale de Lausanne

BM 4138
Station 17
CH-1015, Lausanne


22 June 2013: Our papers "Consistent Discretization of Linear Inverse Problems using Sparse Stochastic Processes" and "Sparse Image Deconvolution with Message Passing" will be presented at SPARS 2013 on July 8 and 11, respectively.

26 May 2013: Unfortunately, I won't be able to attend ICASSP 2013 due to delays in issuance of the Canadian visa.

15 April 2013: The paper "MAP Estimators for Self-Similar Sparse Stochastic Models" was accepted to SAMPTA 2013. It was co-authored with Emrah Bostan, Julien Fageot, and Prof. Michael Unser.

About Me

I am a PhD student in Biomedical Imaging Group (BIG) at EPFL. My supervisor is Prof. Michael Unser. I am working on development of models and algorithms for solving inverse problems.

Currently, I am visiting Information Systems Laboratory (ISL) at Stanford University, where my supervisor is Prof. Andrea Montanari. I am working on statistical inference on random graphs.

Before starting my PhD, I was a visiting student with the Research Lab of Electronics at MIT. I received my MSc and BSc degrees in 2011 and 2008, respectively, from the communication systems department of EPFL. In 2009, I worked as an intern Microsoft in the Unified Communication Group, where I worked on real-time communication platform Microsoft Lync. In 2008, I worked as a research intern in the Signal and Information processing unit of the Telecommunication Research Center in Vienna, Austria.

My full CV can be found [here]. For my LinkedIn go here View Ulugbek Kamilov's profile on LinkedIn. For my Google Scholar profile go [here].

Research Interests

I develop statistical models and algorithms for the reconstruction of biomedical images from highly undersampled or distorted measurements. Such problems are common in various image acquisition modalities such as magnitic resonance imaging (MRI), bio-microscopy, astronomy, etc.

I am interested in the development of iterative algorithms that allow to perform inference on large dimensional data in an efficient way. In particular, message-passing algorithms on graphical models are of particularly high interest to me.

The full list of my publications is available on the right.