“…MMAP estimation consists of minimizing the marginal posterior likelihood p(q | y) after marginalizing out the signal coefficients r [7]. It is a natural estimator for applications where the support may bear more interest than the amplitudes, e.g., for source localization of brain activity using Magneto/ElectroEncephalography (M/EEG) [8], [9], [10].…”
We focus on recovering the support of sparse signals for sparse inverse problems. Using a Bernoulli-Gaussian prior to model sparsity, we propose to estimate the support of the sparse signal using the so-called Marginal Maximum a Posteriori estimate after marginalizing out the values of the nonzero coefficients. To this end, we propose an Expectation-Maximization procedure in which the discrete optimization problem in the Mstep is relaxed into a continuous problem. Empirical assessment with simulated Bernoulli-Gaussian data using magnetoencephalographic lead field matrix shows that this approach outperforms the usual ℓ0 Joint Maximum a Posteriori estimation in Type-I and Type-II error for support recovery, as well as in SNR for signal estimation Index Terms-Sparse coding, inverse problem, Bernoulli-Gaussian model, Marginal-MAP, Joint-MAP.
“…MMAP estimation consists of minimizing the marginal posterior likelihood p(q | y) after marginalizing out the signal coefficients r [7]. It is a natural estimator for applications where the support may bear more interest than the amplitudes, e.g., for source localization of brain activity using Magneto/ElectroEncephalography (M/EEG) [8], [9], [10].…”
We focus on recovering the support of sparse signals for sparse inverse problems. Using a Bernoulli-Gaussian prior to model sparsity, we propose to estimate the support of the sparse signal using the so-called Marginal Maximum a Posteriori estimate after marginalizing out the values of the nonzero coefficients. To this end, we propose an Expectation-Maximization procedure in which the discrete optimization problem in the Mstep is relaxed into a continuous problem. Empirical assessment with simulated Bernoulli-Gaussian data using magnetoencephalographic lead field matrix shows that this approach outperforms the usual ℓ0 Joint Maximum a Posteriori estimation in Type-I and Type-II error for support recovery, as well as in SNR for signal estimation Index Terms-Sparse coding, inverse problem, Bernoulli-Gaussian model, Marginal-MAP, Joint-MAP.
The acoustic modality yields non destructive testing techniques of choice for indepth investigation. Given a precise model of acoustic wave propagation in materials of possibly complex structures, acoustical imaging amounts to the so-called acoustic wave inversion. A less ambitious approach consists in processing pulse-echo data (typically, A-or B-scans) to detect localised echoes with the maximum temporal (and lateral) precision. This is a resolution enhancement problem, and more precisely a sparse deconvolution problem which is naturally addressed in the inversion framework. The paper focuses on the main sparse deconvolution methods and algorithms, with a view to apply them to ultrasonic non-destructive testing.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.