We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Theory

Title: 1-Bit Compressive Sensing via Approximate Message Passing with Built-in Parameter Estimation

Abstract: 1-bit compressive sensing aims to recover sparse signals from quantized 1-bit measurements. Designing efficient approaches that could handle noisy 1-bit measurements is important in a variety of applications. In this paper we use the approximate message passing (AMP) to achieve this goal due to its high computational efficiency and state-of-the-art performance. In AMP the signal of interest is assumed to follow some prior distribution, and its posterior distribution can be computed and used to recover the signal. In practice, the parameters of the prior distributions are often unknown and need to be estimated. Previous works tried to find the parameters that maximize either the measurement likelihood via expectation maximization, which becomes increasingly difficult to solve in cases of complicated probability models. Here we propose to treat the parameters as unknown variables and compute their posteriors via AMP as well, so that the parameters and the signal can be recovered jointly. Compared to previous methods, the proposed approach leads to a simple and elegant parameter estimation scheme, allowing us to directly work with 1-bit quantization noise model. Experimental results show that the proposed approach generally perform much better than the other state-of-the-art methods in the zero-noise and moderate-noise regimes, and outperforms them in most of the cases in the high-noise regime.
Subjects: Information Theory (cs.IT); Signal Processing (eess.SP)
Cite as: arXiv:2007.07679 [cs.IT]
  (or arXiv:2007.07679v2 [cs.IT] for this version)

Submission history

From: Shuai Huang [view email]
[v1] Wed, 15 Jul 2020 13:44:48 GMT (221kb,D)
[v2] Tue, 26 Apr 2022 18:25:25 GMT (167kb,D)

Link back to: arXiv, form interface, contact.