Deep learning for compressive sensing: a ubiquitous systems perspective

Abstract

Compressive sensing (CS) is a mathematically elegant tool for reducing the sensor sampling rate, potentially bringing context-awareness to a wider range of devices. Nevertheless, practical issues with the sampling and reconstruction algorithms prevent further proliferation of CS in real world domains, especially among heterogeneous ubiquitous devices. Deep learning (DL) naturally complements CS for adapting the sampling matrix, reconstructing the signal, and learning from the compressed samples. While the CS–DL integration has received substantial research interest recently, it has not yet been thoroughly surveyed, nor has any light been shed on practical issues towards bringing the CS–DL to real world implementations in the ubiquitous computing domain. In this paper we identify main possible ways in which CS and DL can interplay, extract key ideas for making CS–DL efficient, outline major trends in the CS–DL research space, and derive guidelines for the future evolution of CS–DL within the ubiquitous computing domain.

Multidimensional compressed sensing to advance 23Na multi‐quantum coherences MRI

“…First, further reconstruction improvements could be expected by leveraging convolutional neural networks as proposed by Adlung et al 31 In particular, deep learning techniques could be used to find optimal sparse representations of the signal 55 . Joint frameworks combining Deep Learning and CS have also shown promising results in regards to reduced reconstruction time 56 or overall enhanced reconstruction quality by exploiting relevant features in the images 57 . Nevertheless, the performance highly depends on the amount of training data, which remains limited for 23 Na MQC MRI.…”

“…55 Joint frameworks combining Deep Learning and CS have also shown promising results in regards to reduced reconstruction time 56 or overall enhanced reconstruction quality by exploiting relevant features in the images. 57 Nevertheless, the performance highly depends on the amount of training data, which remains limited for 23 Na MQC MRI.…”

Comparison between a novel compressed sensing-based neural network and traditional neural network approaches for electrical impedance tomography reconstruction

“…The LSTMConvNet model utilized a combination of popular architectures that complement each other for EIT reconstructions [11]. The model was structured in an encoder-decoder style, where an LSTM layer of four units was used to extract important features from the input data.…”

“…The model was structured in an encoder-decoder style, where an LSTM layer of four units was used to extract important features from the input data. The purpose of the LSTM was to extract positional information since the boundary voltage inputs follow a certain order [11]. Such information is often lost when using regular dense layers [12].…”

Reducing ringing artefact in fresnel digital holography using compressed sensing

“…24,25 Recently, machine learning methods have also been proposed for finding the CS reconstruction. 26,27 Next, we outline the OMP algorithm. In a minimum L 1 norm solution using linear programming (LP), we seek:…”