A no-reference hybrid objective QoE evaluation for MPEG-4 encoded video

In order to monitor and ensure adequate Quality of Experience towards end users, this paper proposes a no-reference hybrid objective video quality metrics, based on five high-level parameters extracted from application layer, network layer and bitstream layer. We first cluster various videos based on video content characteristics and then design Back Propagation Neural Network for each content type. The performance evaluation demonstrates a high accuracy in real-time monitoring.