今日论文合集：cs.SD语音17篇，eess.AS音频处理26篇。cs.SD语音【1】 Any-to-Any Generation via Composable Diffusion标题：通过可组合扩散实现任意到任意的一代链接：https://arxiv.org/abs/2305.11846作者：Zineng Tang,Ziyi Yang,Chenguang Zhu,Michael Zeng,Mohit Bansal机构：University of North Carolina at Chapel Hill, Microsoft Azure Cognitive Services Research备注：Project Page: this https URL【2】 Direction Specific Ambisonics Source Separation with End-To-End Deep Learning标题：基于端到端深度学习的方向特定并发源分离链接：https://arxiv.org/abs/2305.11727作者：Francesc Lluís,Nils Meyer-Kahlen,Vasileios Chatziioannou,Alex Hofmann机构：) Dept. of Music Acoustics, University of Music and Performing Arts, Vienna, Austria, ) Dept. for Signal Processing and Acoustics, Aalto University, Otakaari , Espoo, Finland备注：To be published in Acta Acustica. Code and listening examples: this https URL【3】 Sensing of inspiration events from speech: comparison of deep learning and linguistic methods标题：从言语中感知灵感事件：深度学习和语言学方法的比较链接：https://arxiv.org/abs/2305.11683作者：Aki Härmä,Ulf Grossekathöfer,Okke Ouweltjes,Venkata Srikanth Nallanthighal备注：8 pages【4】 MIDI-Draw: Sketching to Control Melody Generation标题：MIDI-DRAW：控制旋律生成的素描链接：https://arxiv.org/abs/2305.11605作者：Tashi Namgyal,Peter Flach,Raul Santos-Rodriguez机构：Department of Computer Science, University of Bristol, Bristol, UK备注：Late-Breaking / Demo Session Extended Abstract, ISMIR 2022 Conference【5】 What You Hear Is What You See: Audio Quality Metrics From Image Quality Metrics标题：您所听到的就是您所看到的：来自图像质量指标的音频质量指标链接：https://arxiv.org/abs/2305.11582作者：Tashi Namgyal,Alexander Hepburn,Raul Santos-Rodriguez,Valero Laparra,Jesus Malo机构：Intelligent Systems Lab, University of Bristol, Bristol, UK, Image Processing Lab, Universitat de Valencia, Valencia, Spain【6】 A Preliminary Study on Augmenting Speech Emotion Recognition using a Diffusion Model标题：利用扩散模型增强语音情感识别的初步研究链接：https://arxiv.org/abs/2305.11413作者：Ibrahim Malik,Siddique Latif,Raja Jurdak,Björn Schuller机构：Emulation AI, Trusted Networks Lab, Queensland University of Technology (QUT), Australia ,GLAM – Group on, Language, Audio, & Music, Imperial College London, UK, Chair of Embedded Intelligence for Health Care and Wellbeing, University of Augsburg, Germany备注：Accepted Interspeech 2023【7】 DUB: Discrete Unit Back-translation for Speech Translation标题：DUB：语音翻译中的离散单元回译链接：https://arxiv.org/abs/2305.11411作者：Dong Zhang,Rong Ye,Tom Ko,Mingxuan Wang,Yaqian Zhou机构：School of Computer Science, Fudan University, ByteDance AI Lab备注：Accepted to Findings of ACL 2023【8】 AlignAtt: Using Attention-based Audio-Translation Alignments as a Guide for Simultaneous Speech Translation标题：AlignAtt：使用基于注意力的音译对齐作为同步语音翻译的指南链接：https://arxiv.org/abs/2305.11408作者：Sara Papi,Marco Turchi,Matteo Negri机构：�Fondazione Bruno Kessler, Italy, ♦University of Trento, Italy, ♮Indipendent Researcher备注：Accepted at Interspeech 2023【9】 Differentially Private Adapters for Parameter Efficient Acoustic Modeling标题：用于参数高效声学建模的差分专用适配器链接：https://arxiv.org/abs/2305.11360作者：Chun-Wei Ho,Chao-Han Huck Yang,Sabato Marco Siniscalchi机构：Georgia Institute of Technology, U.S.A, Kore University of Enna, Italy, Norwegian University of Science and Technology, Norway备注：Accepted to Interspeech 2023. Code will be available at: this https URL The authors would like to express their gratitude to Prof. Chin-Hui Lee from Georgia Tech for providing helpful insights and suggestions【10】 Parameter-Efficient Learning for Text-to-Speech Accent Adaptation标题：文本到语音口音自适应的参数高效学习链接：https://arxiv.org/abs/2305.11320作者：Li-Jen Yang,Chao-Han Huck Yang,Jen-Tzung Chien机构：⋆National Yang Ming Chiao Tung University, Taiwan, †Georgia Institute of Technology, USA备注：Accepted to Interspeech 2023【11】 AMII: Adaptive Multimodal Inter-personal and Intra-personal Model for Adapted Behavior Synthesis标题：AMII：适应行为合成的自适应多模式人与人之间和人内模型链接：https://arxiv.org/abs/2305.11310作者：Jieyeon Woo,Mireille Fares,Catherine Pelachaud,Catherine Achard机构：France, CNRS - ISIR - Sorbonne University备注：8 pages, 1 figure【12】 TrustSER: On the Trustworthiness of Fine-tuning Pre-trained Speech Embeddings For Speech Emotion Recognition标题：TrustSER：用于语音情感识别的精调预训练语音嵌入的可信性链接：https://arxiv.org/abs/2305.11229作者：Tiantian Feng,Rajat Hebbar,Shrikanth Narayanan机构：Signal Analysis and Interpretation Laboratory, Univeristy of Southern California, USA【13】 Pengi: An Audio Language Model for Audio Tasks标题：Pengi：一种面向音频任务的音频语言模型链接：https://arxiv.org/abs/2305.11834作者：Soham Deshmukh,Benjamin Elizalde,Rita Singh,Huaming Wang机构：Microsoft, Carnegie Mellon University备注：Under review【14】 Language-universal phonetic encoder for low-resource speech recognition标题：用于低资源语音识别的语言通用语音编码器链接：https://arxiv.org/abs/2305.11576作者：Siyuan Feng,Ming Tu,Rui Xia,Chuanzeng Huang,Yuxuan Wang机构：Speech and Music Intelligence (SAMI), ByteDance备注：Accepted for publication in INTERSPEECH 2023【15】 Language-Universal Phonetic Representation in Multilingual Speech Pretraining for Low-Resource Speech Recognition标题：用于低资源语音识别的多语言语音预训练中的语言通用语音表示链接：https://arxiv.org/abs/2305.11569作者：Siyuan Feng,Ming Tu,Rui Xia,Chuanzeng Huang,Yuxuan Wang机构：Speech and Music Intelligence (SAMI), ByteDance备注：Accepted for publication in INTERSPEECH 2023【16】 Syllable Discovery and Cross-Lingual Generalization in a Visually Grounded, Self-Supervised Speech Mode标题：基于视觉的自主语音模式下的音节发现和跨语言泛化链接：https://arxiv.org/abs/2305.11435作者：Puyuan Peng,Shang-Wen Li,Okko Räsänen,Abdelrahman Mohamed,David Harwath机构：Department of Computer Science, The University of Texas at Austin, USA, Meta AI, USA,Unit of Computing Sciences, Tampere University, Finland,Rembrand, USA备注：Interspeech 2023. Code & Model: this https URL【17】 Are Microphone Signals Alone Sufficient for Joint Microphones and Sources Localization?标题：单独的麦克风信号是否足以用于联合麦克风和信号源定位？链接：https://arxiv.org/abs/2305.11397作者：Faxian Cao,Yongqiang Cheng,Adil Mehmood Khan,Zhijing Yang备注：2 figureseess.AS音频处理【1】 North Sámi Dialect Identification with Self-supervised Speech Models标题：基于自监督语音模型的北S密语识别链接：https://arxiv.org/abs/2305.11864作者：Sofoklis Kakouros,Katri Hiovain-Asikainen机构：University of Helsinki, Finland, UiT The Arctic University of Norway备注：Accepted at Interspeech 2023【2】 Pengi: An Audio Language Model for Audio Tasks标题：Pengi：一种面向音频任务的音频语言模型链接：https://arxiv.org/abs/2305.11834作者：Soham Deshmukh,Benjamin Elizalde,Rita Singh,Huaming Wang机构：Microsoft, Carnegie Mellon University备注：Under review【3】 Recycle-and-Distill: Universal Compression Strategy for Transformer-based Speech SSL Models with Attention Map Reusing and Masking Distillation标题：循环提取：基于注意图重用和掩蔽提取的Transformer型语音SSL模型的通用压缩策略链接：https://arxiv.org/abs/2305.11685作者：Kangwook Jang,Sungnyun Kim,Se-Young Yun,Hoirin Kim机构：School of Electrical Engineering, KAIST, Graduate School of AI, KAIST备注：Interspeech 2023. Code URL: this https URL【4】 Language-universal phonetic encoder for low-resource speech recognition标题：用于低资源语音识别的语言通用语音编码器链接：https://arxiv.org/abs/2305.11576作者：Siyuan Feng,Ming Tu,Rui Xia,Chuanzeng Huang,Yuxuan Wang机构：Speech and Music Intelligence (SAMI), ByteDance备注：Accepted for publication in INTERSPEECH 2023【5】 BAT: Boundary aware transducer for memory-efficient and low-latency ASR标题：BAT：用于高效内存和低延迟ASR的边界感知传感器链接：https://arxiv.org/abs/2305.11571作者：Keyu An,Xian Shi,Shiliang Zhang机构：Speech Lab of DAMO Academy, Alibaba Group, China备注：accepted into INTERSPEECH2023【6】 Language-Universal Phonetic Representation in Multilingual Speech Pretraining for Low-Resource Speech Recognition标题：用于低资源语音识别的多语言语音预训练中的语言通用语音表示链接：https://arxiv.org/abs/2305.11569作者：Siyuan Feng,Ming Tu,Rui Xia,Chuanzeng Huang,Yuxuan Wang机构：Speech and Music Intelligence (SAMI), ByteDance备注：Accepted for publication in INTERSPEECH 2023【7】 Blank-regularized CTC for Frame Skipping in Neural Transducer标题：神经传感器中用于跳帧的空白正则化CTC链接：https://arxiv.org/abs/2305.11558作者：Yifan Yang,Xiaoyu Yang,Liyong Guo,Zengwei Yao,Wei Kang,Fangjun Kuang,Long Lin,Xie Chen,Daniel Povey机构：Xiaomi Corp., Beijing, China, Shanghai Jiao Tong University, Shanghai, China备注：Accepted in INTERSPEECH 2023【8】 Delay-penalized CTC implemented based on Finite State Transducer标题：基于有限状态变送器的延迟惩罚CTC实现链接：https://arxiv.org/abs/2305.11539作者：Zengwei Yao,Wei Kang,Fangjun Kuang,Liyong Guo,Xiaoyu Yang,Yifan Yang,Long Lin,Daniel Povey机构：Xiaomi Corp., Beijing, China备注：Accepted in INTERSPEECH 2023【9】 Syllable Discovery and Cross-Lingual Generalization in a Visually Grounded, Self-Supervised Speech Mode标题：基于视觉的自主语音模式下的音节发现和跨语言泛化链接：https://arxiv.org/abs/2305.11435作者：Puyuan Peng,Shang-Wen Li,Okko Räsänen,Abdelrahman Mohamed,David Harwath机构：Department of Computer Science, The University of Texas at Austin, USA, Meta AI, USA,Unit of Computing Sciences, Tampere University, Finland,Rembrand, USA备注：Interspeech 2023. Code & Model: this https URL【10】 Are Microphone Signals Alone Sufficient for Joint Microphones and Sources Localization?标题：单独的麦克风信号是否足以用于联合麦克风和信号源定位？链接：https://arxiv.org/abs/2305.11397作者：Faxian Cao,Yongqiang Cheng,Adil Mehmood Khan,Zhijing Yang备注：2 figures【11】 Federated learning for secure development of AI models for Parkinson's disease detection using speech from different languages标题：用于安全开发使用不同语言语音检测帕金森病的人工智能模型的联合学习链接：https://arxiv.org/abs/2305.11284作者：Soroosh Tayebi Arasteh,Cristian David Rios-Urrego,Elmar Noeth,Andreas Maier,Seung Hee Yang,Jan Rusz,Juan Rafael Orozco-Arroyave机构：Pattern Recognition Lab, Friedrich-Alexander-Universit¨at Erlangen-N¨urnberg, Erlangen, Germany, Speech & Language Processing Lab, Friedrich-Alexander-Universit¨at Erlangen-N¨urnberg备注：Accepted for INTERSPEECH 2023【12】 Any-to-Any Generation via Composable Diffusion标题：通过可组合扩散实现任意到任意的一代链接：https://arxiv.org/abs/2305.11846作者：Zineng Tang,Ziyi Yang,Chenguang Zhu,Michael Zeng,Mohit Bansal机构：University of North Carolina at Chapel Hill, Microsoft Azure Cognitive Services Research备注：Project Page: this https URL【13】 Direction Specific Ambisonics Source Separation with End-To-End Deep Learning标题：基于端到端深度学习的方向特定并发源分离链接：https://arxiv.org/abs/2305.11727作者：Francesc Lluís,Nils Meyer-Kahlen,Vasileios Chatziioannou,Alex Hofmann机构：) Dept. of Music Acoustics, University of Music and Performing Arts, Vienna, Austria, ) Dept. for Signal Processing and Acoustics, Aalto University, Otakaari , Espoo, Finland备注：To be published in Acta Acustica. Code and listening examples: this https URL【14】 Sensing of inspiration events from speech: comparison of deep learning and linguistic methods标题：从言语中感知灵感事件：深度学习和语言学方法的比较链接：https://arxiv.org/abs/2305.11683作者：Aki Härmä,Ulf Grossekathöfer,Okke Ouweltjes,Venkata Srikanth Nallanthighal备注：8 pages【15】 MIDI-Draw: Sketching to Control Melody Generation标题：MIDI-DRAW：控制旋律生成的素描链接：https://arxiv.org/abs/2305.11605作者：Tashi Namgyal,Peter Flach,Raul Santos-Rodriguez机构：Department of Computer Science, University of Bristol, Bristol, UK备注：Late-Breaking / Demo Session Extended Abstract, ISMIR 2022 Conference【16】 What You Hear Is What You See: Audio Quality Metrics From Image Quality Metrics标题：您所听到的就是您所看到的：来自图像质量指标的音频质量指标链接：https://arxiv.org/abs/2305.11582作者：Tashi Namgyal,Alexander Hepburn,Raul Santos-Rodriguez,Valero Laparra,Jesus Malo机构：Intelligent Systems Lab, University of Bristol, Bristol, UK, Image Processing Lab, Universitat de Valencia, Valencia, Spain【17】 Phonetic and Prosody-aware Self-supervised Learning Approach for Non-native Fluency Scoring标题：非母语流利性评分的语音和韵律感知的自我监督学习方法链接：https://arxiv.org/abs/2305.11438作者：Kaiqi Fu,Shaojun Gao,Shuju Shi,Xiaohai Tian,Wei Li,Zejun Ma机构：ByteDance【18】 A Preliminary Study on Augmenting Speech Emotion Recognition using a Diffusion Model标题：利用扩散模型增强语音情感识别的初步研究链接：https://arxiv.org/abs/2305.11413作者：Ibrahim Malik,Siddique Latif,Raja Jurdak,Björn Schuller机构：Emulation AI, Trusted Networks Lab, Queensland University of Technology (QUT), Australia ,GLAM – Group on, Language, Audio, & Music, Imperial College London, UK, Chair of Embedded Intelligence for Health Care and Wellbeing, University of Augsburg, Germany备注：Accepted Interspeech 2023【19】 DUB: Discrete Unit Back-translation for Speech Translation标题：DUB：语音翻译中的离散单元回译链接：https://arxiv.org/abs/2305.11411作者：Dong Zhang,Rong Ye,Tom Ko,Mingxuan Wang,Yaqian Zhou机构：School of Computer Science, Fudan University, ByteDance AI Lab备注：Accepted to Findings of ACL 2023【20】 AlignAtt: Using Attention-based Audio-Translation Alignments as a Guide for Simultaneous Speech Translation标题：AlignAtt：使用基于注意力的音译对齐作为同步语音翻译的指南链接：https://arxiv.org/abs/2305.11408作者：Sara Papi,Marco Turchi,Matteo Negri机构：�Fondazione Bruno Kessler, Italy, ♦University of Trento, Italy, ♮Indipendent Researcher备注：Accepted at Interspeech 2023【21】 Differentially Private Adapters for Parameter Efficient Acoustic Modeling标题：用于参数高效声学建模的差分专用适配器链接：https://arxiv.org/abs/2305.11360作者：Chun-Wei Ho,Chao-Han Huck Yang,Sabato Marco Siniscalchi机构：Georgia Institute of Technology, U.S.A, Kore University of Enna, Italy, Norwegian University of Science and Technology, Norway备注：Accepted to Interspeech 2023. Code will be available at: this https URL The authors would like to express their gratitude to Prof. Chin-Hui Lee from Georgia Tech for providing helpful insights and suggestions【22】 Parameter-Efficient Learning for Text-to-Speech Accent Adaptation标题：文本到语音口音自适应的参数高效学习链接：https://arxiv.org/abs/2305.11320作者：Li-Jen Yang,Chao-Han Huck Yang,Jen-Tzung Chien机构：⋆National Yang Ming Chiao Tung University, Taiwan, †Georgia Institute of Technology, USA备注：Accepted to Interspeech 2023【23】 AMII: Adaptive Multimodal Inter-personal and Intra-personal Model for Adapted Behavior Synthesis标题：AMII：适应行为合成的自适应多模式人与人之间和人内模型链接：https://arxiv.org/abs/2305.11310作者：Jieyeon Woo,Mireille Fares,Catherine Pelachaud,Catherine Achard机构：France, CNRS - ISIR - Sorbonne University备注：8 pages, 1 figure【24】 A Parameter-Efficient Learning Approach to Arabic Dialect Identification with Pre-Trained General-Purpose Speech Model标题：一种基于预训练通用语音模型的参数高效学习方法链接：https://arxiv.org/abs/2305.11244作者：Srijith Radhakrishnan,Chao-Han Huck Yang,Sumeer Ahmad Khan,Narsis A. Kiani,David Gomez-Cabrero,Jesper N. Tegner机构：King Abdullah University of Science and Technology, Saudi Arabia, Manipal Institute of Technology, India; ,Georgia Institute of Technology, USA; ,SDAIA-KAUST, Center of Excellence in Data Science and Artificial Intelligence, Thuwal , Saudi Arabia备注：Accepted to Interspeech. Code is available at: this https URL under MIT license【25】 TrustSER: On the Trustworthiness of Fine-tuning Pre-trained Speech Embeddings For Speech Emotion Recognition标题：TrustSER：用于语音情感识别的精调预训练语音嵌入的可信性链接：https://arxiv.org/abs/2305.11229作者：Tiantian Feng,Rajat Hebbar,Shrikanth Narayanan机构：Signal Analysis and Interpretation Laboratory, Univeristy of Southern California, USA【26】 X-LLM: Bootstrapping Advanced Large Language Models by Treating Multi-Modalities as Foreign Languages标题：X-LLM：通过将多个情态视为外语来引导高级大语言模型链接：https://arxiv.org/abs/2305.04160作者：Feilong Chen,Minglun Han,Haozhi Zhao,Qingyang Zhang,Jing Shi,Shuang Xu,Bo Xu机构：Institute of Automation, Chinese Academy of Sciences, School of Future Technology,University of Chinese Academy of Sciences机器翻译，仅供参考
每日语音速递
