Visual Quality

Visual quality assessment aims to objectively measure how humans perceive image and video quality, encompassing factors like sharpness, color accuracy, and the presence of distortions. Current research focuses on developing robust and generalizable models, often employing deep learning architectures like transformers and diffusion models, to address various degradation types (e.g., haze, low light, artifacts) and improve the accuracy of quality predictions, often aligning assessments with human perception through subjective studies and large-scale datasets. This field is crucial for advancing numerous applications, including autonomous driving, virtual reality, and multimedia content creation, by enabling automated quality control and enhancement.