语音/音频处理学术速递[9.21]

【1】 A Large-scale Dataset for Audio-Language Representation Learning标题：一种用于视听语言表征学习的大规模数据集作者：Luoyi Sun,Xuenan Xu,Mengyue Wu,Weidi Xie机构：CMIC & ,X-LANCE, Shanghai Jiao Tong University, Shanghai AI Laboratory链接：http://arxiv.org/pdf/2309.11500v1

【2】 Long-Form End-to-End Speech Translation via Latent Alignment Segmentation标题：基于潜在对齐分割的长格式端到端语音翻译作者：Peter Polák,Ondřej Bojar机构：Ondˇrej BojarCharles University备注：This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible链接：http://arxiv.org/pdf/2309.11384v1

【3】 Incremental Blockwise Beam Search for Simultaneous Speech Translation with Controllable Quality-Latency Tradeoff标题：质量-延迟可控的同声翻译增量顺时针波束搜索算法作者：Peter Polák,Brian Yan,Shinji Watanabe,Alex Waibel,Ondřej Bojar机构：Charles University, Czechia, Carnegie Mellon University, USAJournal-ref：Pol'ak, P., Yan, B., Watanabe, S., Waibel, A., Bojar, O. (2023) Incremental Blockwise Beam Search for Simultaneous Speech Translation with Controllable Quality-Latency Tradeoff. Proc. INTERSPEECH 2023, 3979-3983链接：http://arxiv.org/pdf/2309.11379v1【4】 Automatic Bat Call Classification using Transformer Networks标题：基于Transformer网络的蝙蝠呼叫自动分类作者：Frank Fundel,Daniel A. Braun,Sebastian Gottwald机构：Institute of Neural Information Processing, Ulm University备注：Volume 78, December 2023, 102288链接：http://arxiv.org/pdf/2309.11218v1【5】 Investigating Personalization Methods in Text to Music Generation标题：从文本到音乐生成中的个性化方法研究作者：Manos Plitsis,Theodoros Kouzelis,Georgios Paraskevopoulos,Vassilis Katsouros,Yannis Panagakis机构：Institute for Language and Speech Processing, Athena Research Center, Greece, Department of Informatics and Telecommunications, University of Athens, Greece, Archimedes Unit, Athena Research Center, Greece备注：Submitted to ICASSP 2024, Examples at this https URL链接：http://arxiv.org/pdf/2309.11140v1【6】 Towards Joint Modeling of Dialogue Response and Speech Synthesis based on Large Language Model标题：基于大语言模型的对话应答与语音合成联合建模研究作者：Xinyu Zhou,Delong Chen,Yudong Chen机构：Communication University of China, Hong Kong University of Science and Technology链接：http://arxiv.org/pdf/2309.11000v1【7】 Directional Source Separation for Robust Speech Recognition on Smart Glasses标题：智能眼镜稳健语音识别的方向性信源分离作者：Tiantian Feng,Ju Lin,Yiteng Huang,Weipeng He,Kaustubh Kalgaonkar,Niko Moritz,Li Wan,Xin Lei,Ming Sun,Frank Seide机构：University of Southern California, USA, Meta Platforms Inc., USA备注：Submitted to ICASSP 2024链接：http://arxiv.org/pdf/2309.10993v1【8】 Test-Time Training for Speech标题：语音测试时间训练作者：Sri Harsha Dumpala,Chandramouli Sastry,Sageev Oore机构：Vector Institute and Dalhousie University链接：http://arxiv.org/pdf/2309.10930v1【9】 Semi-Autoregressive Streaming ASR With Label Context标题：具有标签上下文的半自回归流ASR作者：Siddhant Arora,George Saon,Shinji Watanabe,Brian Kingsbury机构：Carnegie Mellon University,IBM Research, Yorktown Heights, USA备注：Submitted to ICASSP 2024链接：http://arxiv.org/pdf/2309.10926v1【10】 Efficient Multi-Channel Speech Enhancement with Spherical Harmonics Injection for Directional Encoding标题：用于定向编码的高效球谐注入多通道语音增强作者：Jiahui Pan,Pengjie Shen,Hui Zhang,Xueliang Zhang机构：College of Computer Science, Inner Mongolia University, China备注：arXiv admin note: text overlap with arXiv:2309.10393链接：http://arxiv.org/pdf/2309.10832v1【11】 Leveraging Data Collection and Unsupervised Learning for Code-switched Tunisian Arabic Automatic Speech Recognition标题：利用数据收集和无监督学习实现码切换突尼斯阿拉伯语自动语音识别作者：Ahmed Amine Ben Abdallah,Ata Kabboudi,Amir Kanoun,Salah Zaiem机构：Tunis Business School; ,University of Michigan-Dearborn;, Abshore; ,LTCI, T´el´ecom Paris, Institut Polytechnique de Paris备注：6 pages, submitted to ICASSP 2024链接：http://arxiv.org/pdf/2309.11327v1【12】 Joint Minimum Processing Beamforming and Near-end Listening Enhancement标题：联合最小处理波束形成和近端监听增强作者：Andreas J. Fuglsig,Jesper Jensen,Zheng-Hua Tan,Lars S. Bertelsen,Jens Christian Lindof,Jan Østergaard机构：⋆ RTX AS, Nørresundby, Denmark, †Aalborg University, Aalborg, Denmark, ©, IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including备注：Submitted to ICASSP 2024链接：http://arxiv.org/pdf/2309.11243v1【13】 Speak While You Think: Streaming Speech Synthesis During Text Generation标题：边说边想：文本生成过程中的流语音合成作者：Avihu Dekel,Slava Shechtman,Raul Fernandez,David Haws,Zvi Kons,Ron Hoory机构：IBM Research备注：Under review for ICASSP 2024链接：http://arxiv.org/pdf/2309.11210v1【14】 Deep Complex U-Net with Conformer for Audio-Visual Speech Enhancement标题：用于视听语音增强的带整形器的深度复杂U网作者：Shafique Ahmed,Chia-Wei Chen,Wenze Ren,Chin-Jou Li,Ernie Chu,Jun-Cheng Chen,Amir Hussain,Hsin-Min Wang,Yu Tsao,Jen-Cheng Hou机构：Academia Sinica ,National Tsing Hua University,National Taiwan University ,Edinburgh Napier University链接：http://arxiv.org/pdf/2309.11059v1【15】 Ensembling Multilingual Pre-Trained Models for Predicting Multi-Label Regression Emotion Share from Speech标题：集成多语言预训练模型预测语音多标签回归情感分享作者：Bagus Tris Atmaja,Akira Sasou机构：∗ National Institute of Advanced Industrial Science and Technology, Japan备注：4 pages, 6 tables, accepted in APSIPA-ASC 2023链接：http://arxiv.org/pdf/2309.11014v1【16】 Discrete Audio Representation as an Alternative to Mel-Spectrograms for Speaker and Speech Recognition标题：离散音频表示作为用于说话人和语音识别的Mel谱图的替代方案作者：Krishna C. Puvvada,Nithin Rao Koluguri,Kunal Dhawan,Jagadeesh Balam,Boris Ginsburg机构：NVIDIA, USA备注：Preprint. Submitted to ICASSP 2024链接：http://arxiv.org/pdf/2309.10922v1【17】 End-to-End Speech Recognition Contextualization with Large Language Models标题：使用大型语言模型的端到端语音识别上下文作者：Egor Lakomkin,Chunyang Wu,Yassir Fathullah,Ozlem Kalinli,Michael L. Seltzer,Christian Fuegen机构：Meta AI

链接：http://arxiv.org/pdf/2309.10917v1

【1】 A Neural TTS System with Parallel Prosody Transfer from Unseen Speakers标题：一种隐形说话人并行韵律转移的神经文语转换系统作者：Slava Shechtman,Raul Fernandez机构：IBM Research AIJournal-ref：Proc. INTERSPEECH 2023, 4853-4857 (2023)链接：http://arxiv.org/pdf/2309.11487v1

【2】 Leveraging Data Collection and Unsupervised Learning for Code-switched Tunisian Arabic Automatic Speech Recognition标题：利用数据收集和无监督学习实现码切换突尼斯阿拉伯语自动语音识别作者：Ahmed Amine Ben Abdallah,Ata Kabboudi,Amir Kanoun,Salah Zaiem机构：Tunis Business School; ,University of Michigan-Dearborn;, Abshore; ,LTCI, T´el´ecom Paris, Institut Polytechnique de Paris备注：6 pages, submitted to ICASSP 2024链接：http://arxiv.org/pdf/2309.11327v1

【3】 Joint Minimum Processing Beamforming and Near-end Listening Enhancement标题：联合最小处理波束形成和近端监听增强作者：Andreas J. Fuglsig,Jesper Jensen,Zheng-Hua Tan,Lars S. Bertelsen,Jens Christian Lindof,Jan Østergaard机构：⋆ RTX AS, Nørresundby, Denmark, †Aalborg University, Aalborg, Denmark, ©, IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including备注：Submitted to ICASSP 2024链接：http://arxiv.org/pdf/2309.11243v1【4】 Speak While You Think: Streaming Speech Synthesis During Text Generation标题：边说边想：文本生成过程中的流语音合成作者：Avihu Dekel,Slava Shechtman,Raul Fernandez,David Haws,Zvi Kons,Ron Hoory机构：IBM Research备注：Under review for ICASSP 2024链接：http://arxiv.org/pdf/2309.11210v1【5】 Deep Complex U-Net with Conformer for Audio-Visual Speech Enhancement标题：用于视听语音增强的带整形器的深度复杂U网作者：Shafique Ahmed,Chia-Wei Chen,Wenze Ren,Chin-Jou Li,Ernie Chu,Jun-Cheng Chen,Amir Hussain,Hsin-Min Wang,Yu Tsao,Jen-Cheng Hou机构：Academia Sinica ,National Tsing Hua University,National Taiwan University ,Edinburgh Napier University链接：http://arxiv.org/pdf/2309.11059v1【6】 Ensembling Multilingual Pre-Trained Models for Predicting Multi-Label Regression Emotion Share from Speech标题：集成多语言预训练模型预测语音多标签回归情感分享作者：Bagus Tris Atmaja,Akira Sasou机构：∗ National Institute of Advanced Industrial Science and Technology, Japan备注：4 pages, 6 tables, accepted in APSIPA-ASC 2023链接：http://arxiv.org/pdf/2309.11014v1【7】 Discrete Audio Representation as an Alternative to Mel-Spectrograms for Speaker and Speech Recognition标题：离散音频表示作为用于说话人和语音识别的Mel谱图的替代方案作者：Krishna C. Puvvada,Nithin Rao Koluguri,Kunal Dhawan,Jagadeesh Balam,Boris Ginsburg机构：NVIDIA, USA备注：Preprint. Submitted to ICASSP 2024链接：http://arxiv.org/pdf/2309.10922v1【8】 End-to-End Speech Recognition Contextualization with Large Language Models标题：使用大型语言模型的端到端语音识别上下文作者：Egor Lakomkin,Chunyang Wu,Yassir Fathullah,Ozlem Kalinli,Michael L. Seltzer,Christian Fuegen机构：Meta AI链接：http://arxiv.org/pdf/2309.10917v1【9】 A Large-scale Dataset for Audio-Language Representation Learning标题：一种用于视听语言表征学习的大规模数据集作者：Luoyi Sun,Xuenan Xu,Mengyue Wu,Weidi Xie机构：CMIC & ,X-LANCE, Shanghai Jiao Tong University, Shanghai AI Laboratory链接：http://arxiv.org/pdf/2309.11500v1【10】 AudioFool: Fast, Universal and synchronization-free Cross-Domain Attack on Speech Recognition标题：AudioFool：快速、通用、无同步的语音识别跨域攻击作者：Mohamad Fakih,Rouwaida Kanj,Fadi Kurdahi,Mohammed E. Fouda备注：10 pages, 11 Figures链接：http://arxiv.org/pdf/2309.11462v1

【11】 Long-Form End-to-End Speech Translation via Latent Alignment Segmentation标题：基于潜在对齐分割的长格式端到端语音翻译作者：Peter Polák,Ondřej Bojar机构： Ondˇrej BojarCharles University备注：This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible链接：http://arxiv.org/pdf/2309.11384v1【12】 Incremental Blockwise Beam Search for Simultaneous Speech Translation with Controllable Quality-Latency Tradeoff标题：质量-延迟可控的同声翻译增量顺时针波束搜索算法作者：Peter Polák,Brian Yan,Shinji Watanabe,Alex Waibel,Ondřej Bojar机构：Charles University, Czechia, Carnegie Mellon University, USAJournal-ref：Pol'ak, P., Yan, B., Watanabe, S., Waibel, A., Bojar, O. (2023) Incremental Blockwise Beam Search for Simultaneous Speech Translation with Controllable Quality-Latency Tradeoff. Proc. INTERSPEECH 2023, 3979-3983链接：http://arxiv.org/pdf/2309.11379v1【13】 Automatic Bat Call Classification using Transformer Networks标题：基于Transformer网络的蝙蝠呼叫自动分类作者：Frank Fundel,Daniel A. Braun,Sebastian Gottwald机构：Institute of Neural Information Processing, Ulm University备注：Volume 78, December 2023, 102288链接：http://arxiv.org/pdf/2309.11218v1【14】 Investigating Personalization Methods in Text to Music Generation标题：从文本到音乐生成中的个性化方法研究作者：Manos Plitsis,Theodoros Kouzelis,Georgios Paraskevopoulos,Vassilis Katsouros,Yannis Panagakis机构： Institute for Language and Speech Processing, Athena Research Center, Greece, Department of Informatics and Telecommunications, University of Athens, Greece, Archimedes Unit, Athena Research Center, Greece备注：Submitted to ICASSP 2024, Examples at this https URL链接：http://arxiv.org/pdf/2309.11140v1【15】 Towards Joint Modeling of Dialogue Response and Speech Synthesis based on Large Language Model标题：基于大语言模型的对话应答与语音合成联合建模研究作者：Xinyu Zhou,Delong Chen,Yudong Chen机构：Communication University of China, Hong Kong University of Science and Technology链接：http://arxiv.org/pdf/2309.11000v1【16】 Directional Source Separation for Robust Speech Recognition on Smart Glasses标题：智能眼镜稳健语音识别的方向性信源分离作者：Tiantian Feng,Ju Lin,Yiteng Huang,Weipeng He,Kaustubh Kalgaonkar,Niko Moritz,Li Wan,Xin Lei,Ming Sun,Frank Seide机构： University of Southern California, USA, Meta Platforms Inc., USA备注：Submitted to ICASSP 2024链接：http://arxiv.org/pdf/2309.10993v1【17】 Test-Time Training for Speech标题：语音测试时间训练作者：Sri Harsha Dumpala,Chandramouli Sastry,Sageev Oore机构：Vector Institute and Dalhousie University链接：http://arxiv.org/pdf/2309.10930v1【18】 Semi-Autoregressive Streaming ASR With Label Context标题：具有标签上下文的半自回归流ASR作者：Siddhant Arora,George Saon,Shinji Watanabe,Brian Kingsbury机构：Carnegie Mellon University,IBM Research, Yorktown Heights, USA备注：Submitted to ICASSP 2024链接：http://arxiv.org/pdf/2309.10926v1【19】 Efficient Multi-Channel Speech Enhancement with Spherical Harmonics Injection for Directional Encoding标题：用于定向编码的高效球谐注入多通道语音增强作者：Jiahui Pan,Pengjie Shen,Hui Zhang,Xueliang Zhang机构：College of Computer Science, Inner Mongolia University, China备注：arXiv admin note: text overlap with arXiv:2309.10393链接：http://arxiv.org/pdf/2309.10832v1