Classification robustness to common optical aberrations

TL;DR

This paper introduces OpticsBench, a benchmark for assessing DNN robustness to realistic optical aberrations, and demonstrates that training with optical kernels improves model performance under such conditions.

Contribution

The paper presents a new benchmark, OpticsBench, for evaluating DNN robustness to optical aberrations and proposes OpticsAugment for improving robustness through data augmentation.

Findings

Performance varies significantly with realistic aberrations compared to ideal kernels.

OpticsAugment improves robustness, with a 21.7% points gain on OpticsBench.

Training with optical kernels enhances accuracy under optical distortions.

Abstract

Computer vision using deep neural networks (DNNs) has brought about seminal changes in people's lives. Applications range from automotive, face recognition in the security industry, to industrial process monitoring. In some cases, DNNs infer even in safety-critical situations. Therefore, for practical applications, DNNs have to behave in a robust way to disturbances such as noise, pixelation, or blur. Blur directly impacts the performance of DNNs, which are often approximated as a disk-shaped kernel to model defocus. However, optics suggests that there are different kernel shapes depending on wavelength and location caused by optical aberrations. In practice, as the optical quality of a lens decreases, such aberrations increase. This paper proposes OpticsBench, a benchmark for investigating robustness to realistic, practically relevant optical blur effects. Each corruption represents an optical aberration (coma, astigmatism, spherical, trefoil) derived from Zernike Polynomials. Experiments on ImageNet show that for a variety of different pre-trained DNNs, the performance varies strongly compared to disk-shaped kernels, indicating the necessity of considering realistic image degradations. In addition, we show on ImageNet-100 with OpticsAugment that robustness can be increased by using optical kernels as data augmentation. Compared to a conventionally trained ResNeXt50, training with OpticsAugment achieves an average performance gain of 21.7% points on OpticsBench and 6.8% points on 2D common corruptions.