Speaker：Xiliang Lv（Wuhan University）

Time：2021-07-23 15:00

Location：Conference Room 105 at Experiment Building at Haiyun Campus

Abstract:

In 1-bit compressive sensing (1-bit CS) where a target signal is coded into a binary measurement, one goal is to recover the signal from noisy and quantized samples. Due to the presence of nonlinearity, noise, and sign flips, it is quite challenging to decode from the 1-bit CS. In this talk, we consider a least squares decoder for several different setting: overdetermined system, with L1-penalty, with vardinality constraint, and with low generative intrinsic dimension. For each case, we show with high probability, the least squares solution approximates the signal up to a constant.  Numerical experiments are presented to illustrate the robustness of the proposed model.