RESEARCH COLLABORATION WITH TOKYO INSTITUTE OF TECHNOLOGY - SPEECH PACKET-LOSS CONCEALMENT IN VOIP AND WIFI NETWORKS

这项旅费资助建议将有助Saeed Vaseghi教授于2009年1月中旬至4月中旬前往日本东京工业大学（TITech）的古井贞树教授的国际知名语音实验室进行为期三个月的研究访问。此次访问的目的是促进布鲁内尔大学和TITech之间的合作，并体验和学习日本大学的学术研究精神，组织文化，跨学科研究工作的文化，日本大学和产业之间的研究整合方法以及东道主的研究方向。在与Furui教授的协议中，此次访问的拟议研究工作将集中在当前研究项目的一部分，该研究项目是对IP语音（VOIP）和WIFI电话网络中用于包丢失隐藏的语音识别和合成模型的各种集成方法进行比较分析。在TITech进行的工作将包括多输入多输出（MIMO）架构，用于估计/预测丢失的语音帧和评估插值算法。

This proposal for a travel grant will facilitate a three months research visit in the period of mid January to mid April 2009 by Prof. Saeed Vaseghi to Prof. Sadaoki Furui's internationally-renowned Speech Laboratory at the Tokyo Institute of Technology (TITech), Japan. The purpose of this visit is to foster collaborations between Brunel University and TITech and to experience and learn from the academic research ethos, the organisational culture, the culture of interdisciplinary research work in Japanese universities, the methods of integration of research between university and the industry in Japan and the research directions of the host. In agreement with Prof. Furui, the proposed research work, for this visit, will focus on parts of a current research project on comparative analysis of the various methods of integration of speech recognition and synthesis models for packet-loss concealment in voice over IP (VOIP) and WIFI phone networks. The work conduced at TITech will include Multiple Input Multiple Output (MIMO) architecture for estimation/prediction of lost speech frames and the evaluation of the interpolation algorithms.