Comparative Performance Analysis: Advantage of IAE Over Signal-Only Models

This chapter evaluates the accuracy of the Individuality Assisted Estimation model in Quality of Experience estimation against traditional signal-only models. The chapter compares IAE’s performance with two state-of-the-art instrumental quality estimation models for speech and video in video-telephony conversations. It assesses the estimation accuracy by measuring the difference between users’ QoE ratings and algorithmically derived scores, demonstrating that the inclusion of individuality data enhances estimation accuracy. The results indicate that while the IAE model provides significant improvements in audiovisual and speech quality estimation, it shows marginal or no improvement in video quality and effort estimation. The chapter also explores the effectiveness of using individuality data alone, revealing its substantial contribution to QoE variability. This comprehensive analysis underscores the value of incorporating user individuality into quality estimation models, despite the modest performance gains observed in some areas.