Omnidirectional Image Quality Assessment

Omnidirectional image quality assessment (OIQA) focuses on objectively measuring the perceived quality of 360° images, crucial for applications like virtual and augmented reality. Current research emphasizes developing robust blind/no-reference OIQA methods, often employing deep learning architectures such as vision transformers and multi-sequence networks that account for the unique viewing characteristics of omnidirectional content, including viewer head movements and viewport selection. These advancements aim to improve the user experience in immersive environments by providing objective metrics for evaluating image quality across various distortions and compression techniques, guiding the development of better image acquisition and processing pipelines.