Sparse signal recovery using Markov random fields

• Presented at: Neural Information Processing Systems (NIPS), Vancouver, B.C., Canada, December 8-11, 2008
• Publication date: 2008

Compressive Sensing (CS) combines sampling and compression into a single sub-Nyquist linear measurement process for sparse and compressible signals. In this paper, we extend the theory of CS to include signals that are concisely represented in terms of a graphical model. In particular, we use Markov Random Fields (MRFs) to represent sparse signals whose nonzero coefficients are clustered. Our new model-based reconstruction algorithm, dubbed Lattice Matching Pursuit (LaMP), stably recovers MRF-modeled signals using many fewer measurements and computations than the current state-of-the-art algorithms.

Reference

• EPFL-CONF-151469

Record created on 2010-09-07, modified on 2012-03-20