Real-time motion-based frame estimation in video lossy transmission

Abstract

In movie transmission, video frames are subject to loss due to noise and/or congestion. The loss of video frames could cause a loss of synchronization between the audio and video streams. If not corrected, this cumulative loss can seriously degrade the motion picture's quality beyond viewers' tolerance. We initially study and classify the effect of audio-video de-synchronization. Then, we develop and study motion-based techniques for estimating the lost frames using the existing received frames, without the need for retransmissions or error control information. The estimated frames are injected at their appropriate locations in the movie stream. The objective is to bring back the synchronization within the tolerance level of viewers, while attempting to find a very close estimation to the original frames at a suitable computation cost.