JIVP Series: Evaluating Visual Quality in Immersive Multimedia: Bridging Human Perception and Machine Intelligence

The rapid evolution of immersive multimedia technologies, such as stereoscopic 3D, virtual reality (VR), light field, and point cloud, has introduced new challenges in evaluating perceptual visual quality in immersive environments. This talk explores the dual perspective of visual quality assessment, combining subjective methods, such as user studies and psychophysical experiments, with objective models designed to predict human perception in immersive environments automatically. We will highlight recent advances in AI-driven analysis, e.g., the use of graph network models, to enhance the accuracy and robustness of quality predictions. By integrating human perceptual factors into machine intelligence, we aim to develop more effective frameworks for assessing immersive visual quality and improving perceptual quality. We hope this talk will inspire insightful discussions and help shape future directions in immersive multimedia quality assessment.