Compressed Learning: A Deep Neural Network Approach

TL;DR

This paper introduces a deep learning framework for compressed learning that jointly optimizes sensing and inference, significantly improving image classification accuracy from limited measurements.

Contribution

It presents an end-to-end deep neural network approach for compressed learning, jointly optimizing sensing and inference for enhanced performance.

Findings

Outperforms state-of-the-art in image classification with limited measurements

Achieves 6.46% error on MNIST at 1% sensing rate

Demonstrates the effectiveness of deep learning in compressed sensing tasks

Abstract

Compressed Learning (CL) is a joint signal processing and machine learning framework for inference from a signal, using a small number of measurements obtained by linear projections of the signal. In this paper we present an end-to-end deep learning approach for CL, in which a network composed of fully-connected layers followed by convolutional layers perform the linear sensing and non-linear inference stages. During the training phase, the sensing matrix and the non-linear inference operator are jointly optimized, and the proposed approach outperforms state-of-the-art for the task of image classification. For example, at a sensing rate of 1% (only 8 measurements of 28 X 28 pixels images), the classification error for the MNIST handwritten digits dataset is 6.46% compared to 41.06% with state-of-the-art.