Impact of realistic properties of the point spread function on classification tasks to reveal a possible distribution shift

Patrick Müller, Alexander Braun, Margret Keuper

Keywords: Distribution Shift, MTF, chromatic aberration, corruption, Deep Neural Networks

Abstract:

Image classification is a long-standing task in computer vision with deep neural
networks (DNN) producing excellent results on various challenges. However, they
are required not only to perform highly accurate on benchmarks such as ImageNet,
but also to robustly handle images in adverse conditions, such as modified lighting,
sharpness, weather conditions and image compression. Various benchmarks
aimed to measure robustness show that neural networks perform differently well
under distribution shifts. While datasets such as ImageNet-C model for example
common corruptions such as blur and adverse weather conditions, we argue that
the properties of the optical system and the potentially resulting complex lens blur
are insufficiently well studied in the literature. This study evaluates the impact of
realistic optical corruptions on the ImageNet classification. The proposed complex
corruption kernels are direction and wavelength dependent and include chromatic
aberration, which are all to be expected in realistic scenarios such as autonomous
driving applications. Our experiments on twelve different DNN models show significant
differences of more than 5% in the top1 classification accuracy, when compared
to the model performances on matched ImageNet-C blur kernels.