A Causal Framework for Aligning Image Quality Metrics and Deep Neural Network Robustness
This addresses the need for image quality metrics aligned with DNN robustness, which is crucial for improving reliability in computer vision applications, representing a novel method for a known bottleneck.
The paper tackled the problem that conventional image quality metrics are poor predictors of deep neural network (DNN) performance for image classification, and developed a causal framework to create new metrics that show strong correlation with DNN performance, enabling effective estimation of image quality distributions for vision tasks.
Image quality plays an important role in the performance of deep neural networks (DNNs) that have been widely shown to exhibit sensitivity to changes in imaging conditions. Conventional image quality assessment (IQA) seeks to measure and align quality relative to human perceptual judgments, but we often need a metric that is not only sensitive to imaging conditions but also well-aligned with DNN sensitivities. We first ask whether conventional IQA metrics are also informative of DNN performance. We show theoretically and empirically that conventional IQA metrics are weak predictors of DNN performance for image classification. Using our causal framework, we then develop metrics that exhibit strong correlation with DNN performance, thus enabling us to effectively estimate the quality distribution of large image datasets relative to targeted vision tasks.