Perceptual Evaluation

Perceptual evaluation assesses how humans perceive the quality of various media, including audio, video, and images, aiming to create objective metrics that accurately reflect subjective experiences. Current research focuses on developing and refining these metrics, particularly for complex media like HDR images and 360° video, often employing machine learning models (e.g., support vector machines, neural networks) trained on large-scale datasets of human ratings. This work is crucial for improving the quality of multimedia content and technologies, enabling more accurate assessment of audio-visual synchronization, speech quality, and other perceptual attributes in diverse applications.