Optimal Mapping Function for Predictions of the Subjective Quality Evaluation Using Artificial Intelligence

With the growth of QoE interest, IPTV providers need a method to control QoE. The paper describes the correlation between the results of objective and subjective methods in video quality assessment. The authors proposed the optimal mapping function for predictions of the subjective quality evaluation based on the objective evaluation to determine the perception of the video quality by the human brain. Our model using artificial intelligence, it is based on a neural network which can simulate and predicts the subjective quality of the scene. It also can predict subjective or objective video quality for video sequences defined by spatial, temporal information, which is the critical and key variable of a given scene, and by the qualitative parameters of the scene. The results from the model are verified by comparing predicted video quality using the proposed classifier with the required value. The two most common statistical parameters related to express performance are Pearson's correlation coefficient and Root Mean Square Error.