Substantial research has explored methods to optimize convolutional neural networks (CNNs) for tasks such as image classification and object detection, but research into the image quality drivers of computer vision performance has been limited. Additionally, there are indications that image degradations such as blur and noise affect human visual interpretation and machine interpretation differently. The General Image Quality Equation (GIQE) predicts overhead image quality for human analysis using the the National Image Interpretability Rating Scale (NIIRS), but no such model exists to predict image quality for interpretation by CNNs. Here, we assess the relationship between image quality variables and convolutional neural network (CNN) performance using an information-theoretic framework. Specifically, we examine the impacts of resolution, blur, and noise on CNN performance for models trained with in-distribution and out-of-distribution distortions. We explore the relationships between CNN performance and image information content as measured by standard Shannon entropy and a similar metric based on image gradients. Using two datasets, we observe that while generalization remains a significant challenge for CNNs faced with out-of-distribution image distortions, CNN performance against low visual quality images remains strong with appropriate training, indicating the potential to expand the design trade space for sensors providing data to computer vision systems
|