语音信息处理、语音识别与合成、个性化语音生成与鉴别、迁移学习

   

081104-模式识别与智能系统

语音识别与合成
语音信息安全
人机交互

2015-09--2018-06   中国科学院自动化研究所   博士
2007-09--2010-07   中国社会科学院研究生院   硕士

   

2020-10~现在, 中国科学院自动化研究所, 副研究员
2018-07~2020-10,中国科学院自动化研究所, 助理研究员
2011-09~2014-11,阿里巴巴集团, 资深算法工程师

2024-08-21-今,IEEE信号处理学会语音与语言处理技术委员会IEEE SLTC, TC Member
2023-12-15-今,中国计算机学会语音对话与听觉专委会, 常务委员
2022-05-07-2022-05-27,ICASSP 2022, Session Chair
2021-12-01-2022-09-24,Interspeech 2022, Area Chairs
2021-04-30-2021-06-15,ICASSP 2021, Session Chair
2020-11-23-今,APSIPA SLA Technical Committee, TC member
2020-10-25-2020-10-29,Interspeech 2020, Session Co-Chairs
2019-09-15-2020-10-30,Interspeech 2020, Area Co-Chairs
2019-09-15-2019-09-19,Interspeech 2019, Session Chair
2019-08-22-今,全国人机语音通讯学术会议(NCMMSC)常设机构, 委员
2019-01-01-2020-04-01,APSIPA 2019, Publication Co-Chairs
2018-12-01-2020-02-01,NCMMSC 2019出版委员会, 出版共同主席

语音信息处理
语音交互

   

（1） 第七届北京市发明专利一等奖, 一等奖, 市地级, 2023
（2） 2022年度中国人工智能学会第十二届吴文俊人工智能科学技术奖, 特等奖, 部委级, 2022
（3） 语音顶级会议国际会议ICASSP 2021多说话人多风格音色克隆大赛极少样本赛道, 一等奖, 其他, 2021
（4） 第十九届全国信号处理学术年会最佳论文, , 其他, 2019
（5） 第十三届全国人机语音通讯学术会议最佳学生论文, , 其他, 2019
（6） Intel AIDC Beijing Best Poster Award, 其他, 2018

[1] []. 基于语音大模型的合成语音检测方法及装置. 中国: CN202311655426.6, 2023-12-05.

[2] []. 基于扩散模型的多模态融合音频生成方法及装置. 中国: CN202311144835.X, 2023-09-06.

[3] 陶建华, 王成龙, 易江燕. 特征提取模型的训练方法和融合发音特征的语音鉴别方法. CN: CN116386602A, 2023-07-04.

[4] 陶建华, 晏鑫蕊, 易江燕. 语音溯源取证方法及装置、设备及存储介质. CN115083422A, 2022-09-10.

[5] 陶建华, 马浩鑫, 易江燕. 一种回放边界负样例的增量式生成语音检测方法和系统. CN115249487A, 2022-10-28.

[6] 陶建华, 马浩鑫, 易江燕. 一种回放边界负样例的增量式生成语音检测方法和系统. CN: CN115249487B, 2023-04-14.

[7] 陶建华, 王成龙, 易江燕, 张震, 李鹏, 石瑾, 孙旭东. 语音检测方法及装置、设备及存储介质. CN: CN114937455A, 2022-08-23.

[8] 陶建华, 许乐, 易江燕. 语音鉴别的数据处理方法和装置. CN115083423A, 2022-09-12.

[9] 陶建华, 张帅, 易江燕. 一种融合多模态语义不变性的语音识别文本增强系统. US17/867,937, 2022-07-19.

[10] 陶建华, 张帅, 易江燕. 一种语音识别与语音翻译端到端系统及设备. US17/852,140, 2022-06-28.

[11] 陶建华, 田正坤, 易江燕. 语音识别模型的训练方法、语音识别方法和系统. US17/836,791, 2022-06-09.

[12] 陶建华, 汪涛, 易江燕, 傅睿博. 编辑音频的方法、装置、电子设备及存储介质. US17/737,666, 2022-05-05.

[13] 傅睿博, 陶建华, 易江燕. 语音检测方法及装置、电子设备及存储介质. CN114420100A, 2022-04-29.

[14] 陶建华, 汪涛, 易江燕, 傅睿博. 一种统一的语音合成与语音转换的训练方法和系统. CN114495898A, 2022-05-13.

[15] 傅睿博, 陶建华, 易江燕. 语音对抗样本生成方法及装置、电子设备及存储介质. CN114267363A, 2022-04-01.

[16] 陶建华, 田正坤, 易江燕. 基于层级区分的生成音频检测系统. US17/674,086, 2022-02-17.

[17] 陶建华, 傅睿博, 易江燕. Method for detecting voice splicing points and storage medium. US17/668,074, 2022-02-09.

[18] 陶建华, 王成龙, 易江燕. 语音场景篡改鉴别的方法、装置、电子设备及存储介质. CN: CN114155875A, 2022-03-08.

[19] 陶建华, 傅睿博, 易江燕. 语音拼接点检测方法及存储介质. US17/668,074, 2022-02-09.

[20] 陶建华, 梁山, 聂帅, 易江燕. 篡改音频的检测方法、装置、电子设备及存储介质. US17/667,212, 2022-02-08.

[21] 陶建华, 王诗明, 傅睿博, 易江燕. 一种细粒度韵律建模的语音生成模型、设备及存储介质. CN114093342A, 2022-02-25.

[22] 陶建华, 田正坤, 易江燕. 语音识别方法、系统、电子设备和存储介质. CN: CN113948085B, 2022-03-25.

[23] 陶建华, 田正坤, 易江燕. 语音识别方法、系统、电子设备和存储介质. CN: CN113948085A, 2022-01-18.

[24] 陶建华, 张帅, 易江燕. 一种可定制的中英混合语音识别端到端系统. CN113936641A, 2022-01-14.

[25] 陶建华, 张帅, 易江燕. 一种可定制的中英混合语音识别端到端系统. CN: CN113936641B, 2022-03-25.

[26] 陶建华, 田正坤, 易江燕. 语音识别模型的训练方法、语音识别方法和系统. CN113936647A, 2022-01-14.

[27] 陶建华, 张帅, 易江燕. 一种语音识别与语音翻译端到端系统及设备. CN113920989A, 2022-01-11.

[28] 聂帅, 梁山, 陶建华, 易江燕. 麦克风阵列语音增强的方法、装置、电子设备及存储介质. CN: CN113889137A, 2022-01-04.

[29] 易江燕, 陶建华, 傅睿博, 田正坤. 生成语音的检测方法、装置、电子设备及存储介质. CN113808579A, 2021-12-17.

[30] 易江燕, 陶建华, 傅睿博, 聂帅, 梁山. 生成语音的检测方法、装置、电子设备及存储介质. CN: CN113808579B, 2022-03-08.

[31] 陶建华, 汪涛, 易江燕. 编辑音频的方法、装置、电子设备及存储介质. CN113724686A, 2021-11-13.

[32] 陶建华, 遆敬苗, 易江燕, 傅睿博. 语音判别方法、装置、电子设备及存储介质. CN: CN113724693A, 2021-11-30.

[33] 傅睿博, 陶建华, 易江燕. 语音拼接点检测方法及存储介质. CN113555007A, 2021-10-26.

[34] 易江燕, 陶建华, 田正坤, 傅睿博. 篡改音频的篡改区域检测方法、装置及存储介质. CN113555037A, 2021-10-26.

[35] 聂帅, 陶建华, 梁山, 易江燕, 傅睿博. 一种语音鉴伪与说话人识别联合建模的方法. CN: CN113555023A, 2021-10-26.

[36] 梁山, 陶建华, 聂帅, 易江燕. 篡改音频的检测方法、装置、电子设备及存储介质. CN: CN113488070A, 2021-10-08.

[37] 陶建华, 马浩鑫, 易江燕. 一种层级分类的生成音频溯源方法及存储介质、计算机设备. CN: CN113488027A, 2021-10-08.

[38] 梁山, 陶建华, 聂帅, 易江燕. 篡改音频的检测方法、装置及存储介质. CN: CN113421592A, 2021-09-21.

[39] 易江燕, 陶建华, 田正坤, 傅睿博. 基于知识迁移的电话信道虚假语音鉴别方法及存储介质. CN113380235A, 2021-09-10.

[40] 易江燕, 陶建华, 田正坤, 傅睿博. 基于知识迁移的电话信道虚假语音鉴别方法及存储介质. CN: CN113380235A, 2021-09-10.

[41] 易江燕, 陶建华, 田正坤, 傅睿博. 一种融合组合模型信息的语音鉴别模型压缩方法. CN113362814A, 2021-09-07.

[42] 陶建华, 马浩鑫, 易江燕. 基于原始波形的轻量级神经网络生成语音鉴别方法和系统. CN: CN113314148A, 2021-08-27.

[43] 陶建华, 马浩鑫, 易江燕. 一种无需原始数据存储的持续性学习生成语音特征的方法. CN113299315A, 2022-08-24.

[44] 易江燕. 基于音素时长特征的虚假语音检测方法及装置. ZL 202110841276.2, 2021-07-26.

[45] 易江燕, 陶建华, 田正坤, 傅睿博. 一种环境对抗的鲁棒语音鉴别方法. CN113284486A, 2021-08-20.

[46] 陶建华, 田正坤, 易江燕. 基于层级区分的生成音频检测系统. CN113284508A, 2021-08-20.

[47] 陶建华, 田正坤, 易江燕. 一种融合多模态语义不变性的语音识别文本增强系统. CN113270086A, 2021-08-17.

[48] 陶建华, 张帅, 易江燕. 统一中英混合文本生成和语音识别的端到端框架. CN: CN113284485A, 2021-08-20.

[49] 陶建华, 张帅, 易江燕. 统一中英混合文本生成和语音识别的端到端系统. CN113284485B, 2021-11-09.

[50] 梁山, 聂帅, 陶建华, 易江燕. 基于相位偏移检测的数字音频篡改取证方法. CN: CN113178199A, 2021-07-27.

[51] 陶建华, 田正坤, 易江燕. 一种流式和非流式混合语音识别系统及流式语音识别方法. CN113257248A, 2021-08-13.

[52] 陶建华, 张大伟, 王华东, 刘通, 杨国花, 易江燕. 基于力导向的知识图谱可视化方法. CN: CN113254670A, 2021-08-13.

[53] 陶建华, 田正坤, 易江燕. 端到端语音转写模型的训练方法、系统、装置. CN: CN110689879B, 2022-02-25.

[54] 陶建华, 易江燕, 温正棋. 语音识别中的小数据语音声学建模方法. CN: CN108682417A, 2018-10-19.

[55] 陶建华, 易江燕, 温正棋, 倪浩. 基于口音瓶颈特征的声学模型自适应方法. CN: CN106875942A, 2017-06-20.

[56] 陶建华, 易江燕, 温正棋, 刘斌. 语音识别中的正则化口音自适应方法. CN: CN106531157A, 2017-03-22.

   

[1] 许乐, 易江燕, 陶建华, 张晓辉, 温正棋, 梁山, 聂帅, 汪涛. 基于自监督预训练模型浅层特征的语音伪造鉴别. 全国人机语音通讯会议. 2022, 第 2 作者
[2] 王成龙, 易江燕, 陶建华, 白烨, 张晓辉, 章楚源. 融合发音特征的伪造语音检测. 第十七届全国人机语音通讯学术会议（NCMMSC2022）. 2022, 第 2 作者
[3] Yi, Jiangyan, Fu, Ruibo, Tao, Jianhua, Nie, Shuai, Ma, Haoxin, Wang, Chenglong, Wang, Tao, Tian, Zhengkun, Bai, Ye, Fan, Cunhang, Liang, Shan, Wang, Shiming, Zhang, Shuai, Yan, Xinrui, Xu, Le, Wen, Zhengqi, Li, Haizhou. ADD 2022: the First Audio Deep Synthesis Detection Challenge. 2022, 第 1 作者http://arxiv.org/abs/2202.08433.
[4] Wang, Tao, Fu, Ruibo, Yi, Jiangyan, Tao, Jianhua, Wen, Zhengqi. NeuralDPS: Neural Deterministic Plus Stochastic Model With Multiband Excitation for Noise-Controllable Waveform Generation. IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING[J]. 2022, 第 3 作者  通讯作者  30: 865-878, http://dx.doi.org/10.1109/TASLP.2022.3140480.
[5] Tian, Zhengkun, Yi, Jiangyan, Tao, Jianhua, Zhang, Shuai, Wen, Zhengqi. Hybrid Autoregressive and Non-Autoregressive Transformer Models for Speech Recognition. IEEE SIGNAL PROCESSING LETTERS[J]. 2022, 第 2 作者  通讯作者  29: 762-766, http://dx.doi.org/10.1109/LSP.2022.3152128.
[6] Fan, Cunhang, Zhang, Hongmei, Yi, Jiangyan, Lv, Zhao, Tao, Jianhua, Li, Taihao, Pei, Guanxiong, Wu, Xiaopei, Li, Sheng. SpecMNet: Spectrum mend network for monaural speech enhancement. APPLIED ACOUSTICS[J]. 2022, 第 3 作者  通讯作者  194: http://dx.doi.org/10.1016/j.apacoust.2022.108792.
[7] Tao wang, Jiangyan Yi, ruibo fu, Tao, Jianhua. Context-Aware Mask Prediction Network for End-to-End Text-Based Speech Editing. ICASSP[J]. 2022, 第 2 作者
[8] 晏鑫蕊, 易江燕, 陶建华, 傅睿博, 汪涛, 温正棋. 谱参数生成精度对音频鉴伪影响的初探-以基于HifiGAN 声码器的频谱局部扰动方法为例. 2022年第十七届全国人机语音通讯学术会议. 2022, 第 2 作者
[9] Tao Wang, Ruibo Fu, Jiangyan Yi, Zhengqi Wen, Tao, Jianhua. Singing-Tacotron: Global Duration Control Attention and Dynamic Filter for End-to-end Singing Voice Synthesis. Proceedings of the 3rd International on Multimodal Sentiment Analysis Workshop and Challenge（ACMMM2022). 2022, 第 3 作者
[10] 汪涛, 易江燕, 傅睿博, 陶建华, 温正棋. CampNet: Context-Aware Mask Prediction for End-to-End Text-Based Speech Editing. IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING[J]. 2022, 第 2 作者  通讯作者  
[11] Fan, Cunhang, Yi, Jiangyan, Tao, Jianhua, Tian, Zhengkun, Liu, Bin, Wen, Zhengqi. Gated Recurrent Fusion With Joint Training Framework for Robust End-to-End Speech Recognition. IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING[J]. 2021, 第 2 作者  通讯作者  29(29): 198-209, http://dx.doi.org/10.1109/TASLP.2020.3039600.
[12] Wang, Tao, Fu, Ruibo, Yi, Jiangyan, Tao, Jianhua, Wen, Zhengqi, Qiang, Chunyu, Wang, Shiming, IEEE. PROSODY AND VOICE FACTORIZATION FOR FEW-SHOT SPEAKER ADAPTATION IN THE CHALLENGE M2VOC 2021. 2021 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP 2021). 2021, 第 3 作者8603-8607, 
[13] 王成龙, 易江燕. 基于全局-时频注意力网络的语音伪造检测. 计算机研究与发展[J]. 2021, 第 2 作者
[14] Bai, Ye, Yi, Jiangyan, Tao, Jianhua, Tian, Zhengkun, Wen, Zhengqi, Zhang, Shuai. Fast End-to-End Speech Recognition Via Non-Autoregressive Models and Cross-Modal Knowledge Transferring From BERT. IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING[J]. 2021, 第 2 作者  通讯作者  29(29): 1897-1911, 
[15] Yi, Jiangyan, Bai, Ye, Tao, Jianhua, Tian, Zhengkun, Wang, Chenglong, Wang, Tao, Fu, Ruibo. Half-Truth: A Partially Fake Audio Detection Dataset. 2021, 第 1 作者http://arxiv.org/abs/2104.03617.
[16] Fu, Ruibo, Tao, Jianhua, Wen, Zhengqi, Yi, Jiangyan, Wang, Tao, Qiang, Chunyu, IEEE. BI-LEVEL STYLE AND PROSODY DECOUPLING MODELING FOR PERSONALIZED END-TO-END SPEECH SYNTHESIS. 2021 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP 2021). 2021, 第 4 作者6568-6572, 
[17] Zhang, Shuai, Yi, Jiangyan, Tian, Zhengkun, Bai, Ye, Tao, Jianhua, Wen, Zhengqi, IEEE. Decoupling_Pronunciation_and_Language_for_End-to-End_Code-Switching_Automatic_Speech_Recognition. 2021 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP 2021). 2021, 第 2 作者6249-6253, 
[18] Wang, Shiming, Ling, Zhenhua, Fu, Ruibo, Yi, Jiangyan, Tao, Jianhua, IEEE. PATNET : A PHONEME-LEVEL AUTOREGRESSIVE TRANSFORMER NETWORK FOR SPEECH SYNTHESIS. 2021 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP 2021). 2021, 第 4 作者5684-5688, 
[19] Bai, Ye, Yi, Jiangyan, Tao, Jianhua, Wen, Zhengqi, Tian, Zhengkun, Zhang, Shuai. Integrating Knowledge Into End-to-End Speech Recognition From External Text-Only Data. IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING[J]. 2021, 第 2 作者  通讯作者  29: 1340-1351, http://dx.doi.org/10.1109/TASLP.2021.3066274.
[20] Ma Haoxin, Yi Jiangyan, Tao Jianhua, Bai Ye, Tian Zhengkun, Wang Chenglong. Continual Learning for Fake Audio Detection. 2021, 第 2 作者http://arxiv.org/abs/2104.07286.
[21] Zhengkun Tian, Jiangyan Yi, Ye Bai, Jianhua Tao, Shuai Zhang, Zhengqi Wen. FSR: Accelerating the Inference Process of Transducer-Based Models by Applying Fast-Skip Regularization. 2021, 第 2 作者http://arxiv.org/abs/2104.02882.
[22] Tao Wang, Ruibo Fu, Jiangyan Yi, Tao Jianhua. Non-autoregressive End-to-End TTS with Coarse-to-Fine Decoding. INTERSPEECH 2020[J]. 2020, 第 3 作者
[23] Tao Wang, Ruibo Fu, Yi Jiangyan, Tao Jianhua. Spoken Content and Voice Factorization for Few-shot Speaker Adaptation. INTERSPEECH 2020[J]. 2020, 第 3 作者
[24] Yi, Jiangyan, Tao, Jianhua. Focal Loss for Punctuation Prediction. INTERSPEECH[J]. 2020, 第 1 作者
[25] Tian Zhengkun, Yi, Jiangyan, Bai Ye, Tao Jianhua, Zhang Shuai, Wen Zhengqi. Synchronous Transformers for End-to-End Speech Recognition. 2020, 第 2 作者http://arxiv.org/abs/1912.02958.
[26] Fan, Cunhang, Tao, Jianhua, Liu, Bin, Yi, Jiangyan, Wen, Zhengqi, Liu, Xuefei. End-to-End Post-Filter for Speech Separation With Deep Attention Fusion Features. IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING[J]. 2020, 第 4 作者28: 1303-1314, http://dx.doi.org/10.1109/TASLP.2020.2982029.
[27] Fu, Ruibo, Tao, Jianhua, Wen, Zhengqi, Yi, Jiangyan, Wang, Tao, IEEE. FOCUSING ON ATTENTION: PROSODY TRANSFER AND ADAPTATIVE OPTIMIZATION STRATEGY FOR MULTI-SPEAKER END-TO-END SPEECH SYNTHESIS. 2020 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING. 2020, 第 4 作者6709-6713, 
[28] Tian, Zhengkun, Yi, Jiangyan, Bai, Ye, Tao, Jianhua, Zhang, Shuai, Wen, Zhengqi, IEEE. SYNCHRONOUS TRANSFORMERS FOR END-TO-END SPEECH RECOGNITION. 2020 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING. 2020, 第 2 作者7884-7888, 
[29] Tao Jianhua, Tao Wang, ruibo fu, Yi, Jiangyan. Dynamic Speaker Representations Adjustment and Decoder Factorization for Speaker Adaptation in End-to-End Speech Synthesis. INTERSPEECH[J]. 2020, 第 4 作者
[30] 陶建华, 傅睿博, 易江燕, 王成龙, 汪涛. 语音伪造与鉴伪的发展与挑战. 信息安全学报[J]. 2020, 第 3 作者5(2): 28-38, http://jcs.iie.ac.cn/xxaqxb/ch/reader/view_abstract.aspx?file_no=20200204&flag=1.
[31] Bai, Ye, Yi, Jiangyan, Tao, Jianhua, Wen, Zhengqi, Fan, Cunhang. A Public Chinese Dataset for Language Model Adaptation. JOURNAL OF SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY[J]. 2020, 第 2 作者  通讯作者  92(8): 839-851, http://dx.doi.org/10.1007/s11265-019-01482-5.
[32] Fan Cunhang, Tao Jianhua, Liu Bin, Yi Jiangyan, Wen Zhengqi. Joint Training for Simultaneous Speech Denoising and Dereverberation with Deep Embedding Representations. INTERSPEECH[J]. 2020, 第 4 作者
[33] Tao Wang, Tao Jianhua, ruibo fu, Yi, Jiangyan. Bi-level Speaker Supervision for One-shot Speech Synthesis. INTERSPEECH 2020[J]. 2020, 第 4 作者
[34] ruibo fu, Tao Jianhua, Zhengqi Wen, Yi, Jiangyan. Dynamic Soft Windowing and Language Dependent Style Token for Code-Switching End-to-End Speech Synthesis. INTERSPEECH[J]. 2020, 第 4 作者
[35] Fan, Cunhang, Yi, Jiangyan, Tao, Jianhua, Tian, Zhengkun, Liu, Bin, Wen, Zhengqi. Gated Recurrent Fusion with Joint Training Framework for Robust End-to-End Speech Recognition. 2020, 第 2 作者http://arxiv.org/abs/2011.04249.
[36] Ye Bai, Jiangyan Yi, Jianhua Tao, Zhengkun Tian, Zhengqi Wen, Shuai Zhang. Listen Attentively, and Spell Once: Whole Sentence Generation via a Non-Autoregressive Architecture for Low-Latency Speech Recognition. 2020, 第 2 作者http://arxiv.org/abs/2005.04862.
[37] Zhengkun Tian, Jiangyan Yi, Jianhua Tao, Ye Bai, Shuai Zhang, Zhengqi Wen. Spike-Triggered Non-Autoregressive Transformer for End-to-End Speech Recognition. 2020, 第 2 作者http://arxiv.org/abs/2005.07903.
[38] Fan Cunhang, Tao Jianhua, Liu Bin, Yi Jiangyan, Wen Zhengqi. Gated Recurrent Fusion of Spatial and Spectral Features for Multi-channel Speech Separation with Deep Embedding Representations. Interspeech[J]. 2020, 第 4 作者
[39] Yi, Jiangyan, Tao, Jianhua. SELF-ATTENTION BASED MODEL FOR PUNCTUATION PREDICTION USING WORD AND SPEECH EMBEDDINGS. 2019 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP)[J]. 2019, 第 1 作者  通讯作者  7270-7274, 
[40] Yi, Jiangyan, Tao, Jianhua, Wen, Zhengqi, Bai, Ye. Language-Adversarial Transfer Learning for Low-Resource Speech Recognition. IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING[J]. 2019, 第 1 作者27(3): 621-630, http://ir.ia.ac.cn/handle/173211/25289.
[41] 傅睿博, 陶建华, 温正棋, 易江燕, 王诗明, 强春雨. 多语言语音数据库自动优化方法研究. 2019, 第 4 作者
[42] Yi, Jiangyan, Tao, Jianhua, Bai, Ye. LANGUAGE-INVARIANT BOTTLENECK FEATURES FROM ADVERSARIAL END-TO-END ACOUSTIC MODELS FOR LOW RESOURCE SPEECH RECOGNITION. 2019 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP)[J]. 2019, 第 1 作者  通讯作者  6071-6075, 
[43] Fan Cunhang, Liu Bin, Tao Jianhua, Yi Jiangyan, Wen Zhengqi. Discriminative Learning for Monaural Speech Separation Using Deep Embedding Features. interspeech2019. 2019, 第 4 作者http://arxiv.org/abs/1907.09884.
[44] Zheng, Yibin, Tao, Jianhua, Wen, Zhengqi, Yi, Jiangyan. Forward-Backward Decoding Sequence for Regularizing End-to-End TTS. IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING[J]. 2019, 第 4 作者27(12): 2067-2079, http://dx.doi.org/10.1109/TASLP.2019.2935807.
[45] Ye Bai, Jiangyan Yi, Tao Jianhua. A Time Delay Neural Network with Shared Weight Self-Attention for Small-Footprint Keyword Spotting. INTERSPEECH[J]. 2019, 第 2 作者
[46] Ye Bai, Jiangyan Yi, Jianhua Tao, Zhengkun Tian, Zhengqi Wen. Learn Spelling from Teachers: Transferring Knowledge from Language Models to Sequence-to-Sequence Speech Recognition. 2019, 第 2 作者http://arxiv.org/abs/1907.06017.
[47] Zhengkun Tian, Jiangyan Yi, Jianhua Tao, Ye Bai, Zhengqi Wen. Self-Attention Transducers for End-to-End Speech Recognition. 2019, 第 2 作者http://arxiv.org/abs/1909.13037.
[48] Jiangyan Yi, Jianhua Tao, Zhengqi Wen, Ye Bai. Language-Adversarial Transfer Learning for Low-Resource Speech Recognition. IEEE / ACM TRANSACTIONS ON AUDIO, SPEECH AND LANGUAGE PROCESSING (TASLP). 2019, 第 1 作者27(3): http://kns.cnki.net/KCMS/detail/detail.aspx?QueryID=0&CurRec=1&recid=&FileName=SJCM0D850DF295CFFB91AA79EEA520C4B2A5&DbName=WWMERGEJ01&DbCode=WWME&yx=&pr=&URLID=&bsm=.
[49] 范存航, 刘斌, 陶建华, 温正棋, 易江燕. 一种基于卷积神经网络的端到端语音分离方法. 信号处理[J]. 2019, 第 5 作者35(4): 542-548, http://lib.cqvip.com/Qikan/Article/Detail?id=7002150057.
[50] 易江燕, 陶建华, 刘斌, 温正棋. 基于迁移学习的噪声鲁棒语音识别声学建模. 清华大学学报：自然科学版[J]. 2018, 第 1 作者58(1): 55-60, http://lib.cqvip.com/Qikan/Article/Detail?id=674347161.
[51] Yi, Jiangyan, Wen, Zhengqi, Tao, Jianhua, Ni, Hao, Liu, Bin. CTC Regularized Model Adaptation for Improving LSTM RNN Based Multi-Accent Mandarin Speech Recognition. JOURNAL OF SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY[J]. 2018, 第 1 作者90(7): 985-997, http://dx.doi.org/10.1007/s11265-017-1291-1.
[52] Huang, Jian, Tao, Jianhua, Li, Ya, Lian, Zheng, Yi, Jiangyan. End-to-End Continuous Emotion Recognition from Video Using 3D Convlstm Networks. ICASSP, IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING. 2018, 第 5 作者http://ir.ia.ac.cn/handle/173211/22375.
[53] Yi Jiangyan, Tao Jianhua, Wen Zhengqi, Bai Ye, IEEE. ADVERSARIAL MULTILINGUAL TRAINING FOR LOW-RESOURCE SPEECH RECOGNITION. 2018 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP). 2018, 第 1 作者4899-4903, http://apps.webofknowledge.com/CitedFullRecord.do?product=UA&colName=WOS&SID=5CCFccWmJJRAuMzNPjj&search_mode=CitedFullRecord&isickref=WOS:000446384605014.
[54] Huang,Jian, Tao, Jianhua, Lian, Zheng, Li, Ya, Yi, Jiangyan, Niu, Mingyue. Speech Emotion Recognition Using Semi-supervised Learning with Ladder Networks. 2018 1ST ASIAN CONFERENCE ON AFFECTIVE COMPUTING AND INTELLIGENT INTERACTION, ACII ASIA 2018. 2018, 第 5 作者http://ir.ia.ac.cn/handle/173211/22377.
[55] 易江燕, 陶建华, 刘斌, 温正棋. 基于迁移学习的鲁棒语音识别声学建模方法. 清华大学学报[J]. 2018, 第 1 作者58(1): 55-60, http://ir.ia.ac.cn/handle/173211/19898.
[56] Yi Jiangyan, NiHao, Tao Jianhua, Wen Zhengqi. Acoustic Model Compression with Knowledge Transfer. 2017, 第 1 作者http://ir.ia.ac.cn/handle/173211/19888.
[57] Jiangyan Yi, Tao, Jianhua, Zhengqi Wen, Hao Ni. Distilling Knowledge from an Ensemble of Models for Punctuation Prediction. INTERSPEECH 2017[J]. 2017, 第 1 作者
[58] 乌日其其格, 陶建华, 白音门德, 易江燕, 温正棋, 白烨. 面向语音识别的蒙古语标准音语音库的建立. 2017, 第 4 作者http://ir.ia.ac.cn/handle/173211/19916.
[59] Bai Ye, Yi Jiangyan, Ni Hao, Wen Zhengqi, Liu Bin, Li Ya, Tao Jianhua. End-to-end Keywords Spotting Based on Connectionist Temporal Classification for Mandarin. 2016, 第 2 作者http://ir.ia.ac.cn/handle/173211/19742.
[60] Yi Jiangyan, Tao Jianhua, Ni Hao, Wen Zhengqi. Improving BLSTM RNN Based Mandarin Speech Recognition Using Accent Dependent Bottleneck Features. 2016, 第 1 作者http://ir.ia.ac.cn/handle/173211/19881.
[61] Yi, Jiangyan, Ni, Hao, Wen, Zhengqi, Tao, Jianhua, IEEE. Improving BLSTM RNN Based Mandarin Speech Recognition Using Accent Dependent Bottleneck Features. 2016 ASIA-PACIFIC SIGNAL AND INFORMATION PROCESSING ASSOCIATION ANNUAL SUMMIT AND CONFERENCE (APSIPA). 2016, 第 1 作者
[62] Ni Hao, Yi Jiangyan, Wen Zhengqi, Tao Jianhua, Tan T, Li X, Chen X, Zhou J, Yang J, Cheng H. Recurrent Neural Network Based Language Model Adaptation for Accent Mandarin Speech. PATTERN RECOGNITION (CCPR 2016), PT II. 2016, 第 2 作者663: 607-617, 
[63] Yi, Jiangyan, Ni, Hao, Wen, Zhengqi, Liu, Bin, Tao, Jianhua. CTC Regularized Model Adaptation for Improving LSTM RNN Based MultiAccent Mandarin Speech Recognition. 2016 10TH INTERNATIONAL SYMPOSIUM ON CHINESE SPOKEN LANGUAGE PROCESSING (ISCSLP)[J]. 2016, 第 1 作者http://apps.webofknowledge.com/CitedFullRecord.do?product=UA&colName=WOS&SID=5CCFccWmJJRAuMzNPjj&search_mode=CitedFullRecord&isickref=WOS:000405610900058.
[64] Bai, Ye, Yi, Jiangyan, Ni, Hao, Wen, Zhengqi, Liu, Bin, Li, Ya, Tao, Jianhua. End-to-end Keywords Spotting Based on Connectionist Temporal Classification for Mandarin. 2016 10TH INTERNATIONAL SYMPOSIUM ON CHINESE SPOKEN LANGUAGE PROCESSING (ISCSLP)[J]. 2016, 第 2 作者http://apps.webofknowledge.com/CitedFullRecord.do?product=UA&colName=WOS&SID=5CCFccWmJJRAuMzNPjj&search_mode=CitedFullRecord&isickref=WOS:000405610900098.

   

（ 1 ） 面向小数据语音建模的跨语言迁移学习研究, 负责人, 国家任务, 2020-01--2022-12
（ 2 ） 基于小数据的强噪声语音识别声学模型研究, 负责人, 中国科学院计划, 2019-01--2021-12
（ 3 ） 网络多媒体鉴伪关键技术研究, 负责人, 国家任务, 2020-07--2023-06
（ 4 ） 语音识别及语音合成技术研究合作项目, 负责人, 境内委托项目, 2021-03--2022-03
（ 5 ） 大数据多模态交互协同关键技术, 参与, 国家任务, 2017-10--2021-09
（ 6 ） 大数据分析关键技术及应用, 参与, 中国科学院计划, 2020-01--2020-12
（ 7 ） 基于多语种多通道融合的XXX, 参与, 国家任务, 2018-02--2021-12
（ 8 ） 语音关键技术, 负责人, 境内委托项目, 2021-06--2022-09
（ 9 ） 基于连续学习的音视频分析关键技术, 负责人, 国家任务, 2021-10--2024-09

现指导学生

顾浩  硕士研究生  081104-模式识别与智能系统  

宋瀚林  博士研究生  081203-计算机应用技术  

张豪  博士研究生  081104-模式识别与智能系统  

任勇  博士研究生  081104-模式识别与智能系统  

周俊佐  硕士研究生  085410-人工智能  

张一诺  硕士研究生  085400-电子信息  

马一鸣  硕士研究生  085400-电子信息  

程振华  硕士研究生  085400-电子信息  

曾思丁  硕士研究生  085410-人工智能  

徐涛  硕士研究生  085410-人工智能  

杨龙江  硕士研究生  085410-人工智能