Does deep learning always outperform simple linear regression in optical imaging?

TL;DR

This paper compares deep learning and linear regression in optical imaging, showing that linear methods can outperform deep learning especially with limited training data, due to the linear nature of many optical systems.

Contribution

The study reveals that simple linear regression can outperform deep learning in certain optical imaging tasks, challenging the assumption that deep learning is always superior.

Findings

Linear regression can outperform deep learning with small training datasets.

Deep learning's weakness is pronounced when training data is limited.

Many optical systems are linear, making simple methods more suitable.

Abstract

Deep learning has been extensively applied in many optical imaging applications in recent years. Despite the success, the limitations and drawbacks of deep learning in optical imaging have been seldom investigated. In this work, we show that conventional linear-regression-based methods can outperform the previously proposed deep learning approaches for two black-box optical imaging problems in some extent. Deep learning demonstrates its weakness especially when the number of training samples is small. The advantages and disadvantages of linear-regression-based methods and deep learning are analyzed and compared. Since many optical systems are essentially linear, a deep learning network containing many nonlinearity functions sometimes may not be the most suitable option.