人物生平
2012年7月-2015年10月,中國科學院自動化研究所,模式識別國家重點實驗室,助理研究員
2014年5月-2014年9月,
愛爾蘭聖三一學院,計算機與統計學院,Research Fellow
2015年11月-2021年,中國科學院自動化研究所,模式識別國家重點實驗室,副研究員
研究領域
智慧型人機語音互動技術是一種簡單、便捷的人機互動方式。語音互動技術包括了語音識別、情感識別、語義理解、語音合成等技術。隨著大數據和計算技術的快速發展,特別是深度神經網路的發展,
智慧型語音互動技術取得了一系列的重大突破,為智慧型語音互動技術走向實用化提供了可能。
主要研究領域包括語音合成、多模態情感識別和自然人機對話技術。
社會兼職
全國人機語音通訊學術會議常設機構,委員,2015-
NCMMSC2015,情感計算特殊議題,共同主席,2015
Program committee member of ISCSLP, 2014-
Sponsorship Chair of International Conference on Affective Computing and Intelligent Interaction 2015, 2015-
Program Chair of AMAI workshop, 2015
Program Chair of Multimodal Emotion Recognition Challenge, 2016-
專利與獎勵
專利
(1)一種對國語重音進行層次化建模和預測的方法,ZL201110200330.1,陶建華,李雅
獎勵
(1)“具有個性化自適應能力的高性能語音處理技術及套用”,2014年北京市科學技術獎二等獎,第二完成人
(2)指導研究生曾獲頂級會議INTERSPEECH 2016的最佳學生論文。第三完成人
(3)AVEC2014,AVEC2015國際維度情感識別競賽第二名。
(4)“採用重音調整模型的HMM語音合成系統”,2011年全國人機語音通訊學術會議最佳學生論文提名獎,第一完成人。
科研活動
(1)語音評價中的韻律建模和評價方法研究,自然基金,2014-2016,負責人
(2)基於維度模型的情感語音建模及生成方法研究,自然基金,2013-2015, 骨幹
(3)面向移動終端的多模態自然互動技術,863項目,2015-2017,課題骨幹
(4)社會情感的語音生成與認知的跨語言跨文化研究,社科重點,2014-2018,參與
(5)基於跨語言韻律模型的自適應語音合成,2010-2012,自然基金,參與
合作情況
與百度、三星、東芝,騰訊等公司在語音技術方面展開多次合作。
出版信息
[1].Ya Li, Jianhua Tao, Wei Lai, Xiaoying Xu, Quantitative intonation modeling of interrogative sentences for Mandarin Speech Synthesis, Speech communication, 2017.
[2].Ya Li, Jianhua Tao, Linlin Chao,Wei Bao, Yazhu Liu, CHEAVD: a Chinese natural emotional audio–visual database, Journal of Ambient Intelligence and Humanized Computing, 2016, DOI: 10.1007/s12652-016-0406-z.
[3].Ya Li, Jianhua Tao, Keikichi Hirose, Xiaoying Xu, Wei Lai, Hierarchical stress modeling and generation in mandarin for expressive Text-to-Speech, Speech communication, 2015, Vol. 72,pp.59-73.
[4].Hao Che,Ya Li*,Jianhua Tao,Zhengqi Wen, "Investigating Effect of Rich Syntactic Features on Mandarin Prosodic Phrase Boundaries Prediction" Journal of Signal Processing Systems,2016, 82(2):263-271
[5].Wei Lai, Jiahong Yuan, Ya Li, Xiaoying Xu, Mark Liberman, The rhythmic constraint of prosodic boundaries in Chinese Mandarin based on corpora of silent reading and speech perception, INTERSPEECH 2016, pp.87-91. 最佳學生論文
[6].Yibin Zheng, Ya Li, Zhengqi Wen, Xingguang Ding, Jianhua Tao, ”Improving Prosodic Boundaries Prediction for Mandarin Speech Synthesis by Using Enhanced Embedding Feature and Model Fusion Approaches”, Interspeech 2016, PP:3201-3205, Sep 8-12, 2016.
[7].Ya Li, Tao J, Schuller B, Shan S, Jiang D, Jia J (2016) MEC 2016: The multimodal emotion recognition challenge of CCPR 2016. In: Chinese Conference on Pattern Recognition (CCPR), Chengdu, China, pp. 667-678.
[8].Zhengqi Wen, Ya Li and Jianhua Tao, “The Parameterized Phoneme Identity Feature as a Continuous Real-Valued Vector for Neural Network based Speech Synthesis”, Interspeech 2016, PP:2248-2252, Sep 8-12, 2016.
[9].Yibin Zheng, Ya Li, Zhengqi Wen, Bin Liu, Jianhua Tao, ”Investigating Deep Neural Network Adaptation for Generating Exclamatory and Interrogative Speech in Mandarin”, 10th International Symposium on Chinese Spoken Language Processing (ISCSLP2016), Oct 17-20, 2016.
[10].Yibin Zheng, Ya Li, Zhengqi Wen, Bin Liu, Jianhua Tao, “Text-based sentential stress prediction using continuous lexical embedding for Mandarin speech synthesis”, 10th International Symposium on Chinese Spoken Language Processing (ISCSLP2016), Oct 17-20, 2016.
[11].Ya Li, Nick Campbell, Jianhua Tao, Voice Quality: Not Only About “You” But Also About “Your Interlocutor”, ICASSP 2015, pp. 4739-4743.
[18].Ya Li, Jianhua Tao, Xiaoying Xu, Hierarchical Stress Modeling in Mandarin Text-to-Speech, InterSpeech 2011, Florence, Italy, 2013-2016.
[19].Ya Li, Jianhua Tao, Meng Zhang, Shifeng Pan, Xiaoying Xu, Text-based unstressed syllable prediction in Mandarin, InterSpeech 2010, 26-30, September, Makuhari, Chiba,Japan,pp.1752-1755
[20].Ya Li, Shifeng Pan, Jianhua Tao, HMM-based Expressive Speech Synthesis with a Flexible Mandarin Stress Adaptation Model, ICSP 2010, Beijing, Oct 24-28.pp.625-628.