本文選題：動態(tài)面部表情　切入點：情緒化　出處：《計算機(jī)應(yīng)用研究》2016年12期

【摘要】：針對現(xiàn)有的虛擬說話人面部表情比較單一,表情和動作不能很好地協(xié)同的問題,提出了一種建立具有真實感的情緒化虛擬人的方法。該方法首先利用三參數(shù)產(chǎn)生,保持和消退來對動態(tài)面部表情進(jìn)行仿真,采用融合變形技術(shù)合成融合多元素復(fù)雜的表情,然后以人類心理學(xué)的統(tǒng)計數(shù)據(jù)為依據(jù)來對眼部和頭部動作進(jìn)行設(shè)計,使虛擬人看起來更加逼真,最后分析了外在條件相機(jī)位置、光照對增加虛擬人真實感的影響。實驗結(jié)果表明,該方法建立的虛擬人不僅逼真自然且富有感情,而且語音、動態(tài)面部表情、眼動和頭動達(dá)到了很好的協(xié)調(diào)同步。
[Abstract]:In order to solve the problem that the facial expression of virtual speaker is single and the facial expression and action can not be well coordinated, a method of establishing emotional virtual human with realistic feeling is proposed. The method is first generated with three parameters. Maintain and fade to simulate the dynamic facial expression, use the fusion deformation technology to synthesize the fusion multi-element complex facial expression, then based on the statistical data of human psychology to design the eye and head movement. Finally, the effect of external camera position and illumination on increasing virtual human's reality is analyzed. The experimental results show that the proposed method is not only realistic, natural and emotional, but also has voice. Dynamic facial expressions, eye movements and head movements achieve good coordination and synchronization.
【作者單位】： 西南交通大學(xué)信息科學(xué)與技術(shù)學(xué)院;
【基金】：成都市科技資助項目(科技惠民技術(shù)研發(fā)項目)(2015-HM01-00050-SF) 四川省動漫研究中心2015年度科研資助項目(DM201504) 國家自然科學(xué)基金面上資助項目(61371165)
【分類號】：TP391.41
，

本文編號：1675661