Content-Based Prediction of Optimal Video Adaptation Operations Using Subjective Quality Evaluation

Video adaptation allows for direct manipulation of existing encoded video streams to meet new resource constraints without having to encode the video from scratch. Multi-dimensional scalable coding such as motion-compensated subband coding (MCSBC) offers an effective and flexible representation for video adaptation, which is important in a universal media access scenario. However, currently most of the adaptation research is focused on the issue of predefined video adaptation method and there is little work on systematical solution in selecting optimal operation among multi-dimensional adaptations. In this paper we utilize a content-based prediction framework to select optimal spatio-temporal video adaptation operation using subjective quality evaluation. Content features in the compress domain are employed to explore the correlation between the video content characteristic and optimal adaptation behavior during a pattern classification procedure. The experiment result indicates that our proposed method can efficiently predict the right adaptation operation matching subjective quality evaluation.