| With the increased needs of audio and video entertainment, TelepresenceCommunication has come into daily life. And the audio has changed a lot from monochannel to multi-chennel audio such as22.2channel even more channels so as to enhancehuman’s spatial experience.Howeverwhen the channels are large, the bandwidth becomes alimited factor.So thetransmission of stereo and multi-channel is a key problem.How to codethe audio in less bit rates and to get the audio with better quality is our major concern.G.719(G.722.1-FB), which is a low complexity full-band audio codec, mainly appliedin interactive communication for example videoconference,and which is proposed byITU-T in the year of2008. The coding algorithm is based on transform coding withadaptive time-resolution, adaptive bit-allocation and low complexity lattice vectorquantization.This thesis introduces several common stereo and multi-channel processingtechnologiesin details. Considering the needs of HD voice and audio communication, weintroduce stereo and multi-channel coding model to expand its applications. This paperproposed stereo coding method in G.719codec based on the CELT layer in Opus codec.Besides, in view of spatial parameter coding could reduce itsstorage space and bandwidth,we proposes parameter stereo and multi-cahnnel coding method in G.719codec. Theexperimental results indicated that the parametric stereo and multi-channel method canobtain a high coding gain.Besides, in recent years many researchers have got achievements in the field of tensordata mining, and applied it to the field of medicine,vedio image and traffic data restore.Inthis paper, a method based on CP decomposition weighted optimization algorithm toreteievel5.1multi-channel audio signal with missing data is proposed.We also adopt theMushra method to test the performance of the proposed scheme.The experimental resultsshow that the quality of retrieval audio is good. |
PDF Full Text Request