共有仮想空間アバタ音声チャットシステムにおける音声情報を利用した表情制御(コミュニケーション支援及び一般)

概要

論文の詳細を見る
An automatic smile-level control algorithm of CG avatar based on the user's utterance is proposed, in order to facilitate multi-party casual chat in a virtual-space voice chat system. The proposed algorithm is comprised of three components those are calculated from the user and partner's utterance. The first is the smile-level continuous component SLr that is defined as the moving-avarage of the normalized fundamental frequency of the partner's voice, to reflect the mood. The second is the smile-level rapid component SLr that is calculated from the initial voice volume of the partner, to reflect the rapid emotion change caused by the user's utterance. The third is the smile-level pseudo intention expression component SLpi that is calculated from the final phase voice volume of the user, to simulate the social response. A prototype virtual-space voice chat system demonstrated the effectiveness of the proposed algorithm.
社団法人電子情報通信学会の論文
2007-05-17