Super Resolution Image Quality Assessment

Super-resolution image quality assessment (SR-IQA) focuses on developing objective methods to evaluate the visual quality of images upscaled by super-resolution algorithms, a crucial task given the increasing prevalence of SR technology. Current research emphasizes the development of both full-reference and no-reference/reduced-reference IQA methods, often employing deep learning architectures like convolutional neural networks and vision transformers, incorporating factors such as scale information and balancing perceptual quality with structural fidelity. Improved SR-IQA metrics are vital for advancing SR algorithm development and ensuring the reliable assessment of image quality in various applications, from broadcast media to medical imaging.