今日论文合集：cs.SD语音22篇，eess.AS音频处理27篇。cs.SD语音【1】 Q&A: Query-Based Representation Learning for Multi-Track Symbolic Music re-Arrangement标题：问答：基于查询的多声道符号音乐重排表示学习链接：https://arxiv.org/abs/2306.01635作者：Jingwei Zhao,Gus Xia,Ye Wang机构：School of Computing, NUS, Institute of Data Science, NUS, Integrative Sciences and Engineering Programme, NUS Graduate School, Music X Lab, NYU Shanghai, MBZUAI备注：Accepted by IJCAI 2023 Special Track for AI the Arts and Creativity【2】 Enhance Temporal Relations in Audio Captioning with Sound Event Detection标题：利用声音事件检测增强音频字幕中的时间关系链接：https://arxiv.org/abs/2306.01533作者：Zeyu Xie,Xuenan Xu,Mengyue Wu,Kai Yu机构：MoE Key Lab of Artificial Intelligence X-LANCE Lab, Department of Computer Science and Engineering AI Institute, Shanghai Jiao Tong University, Shanghai, China【3】 Learning Local to Global Feature Aggregation for Speech Emotion Recognition标题：用于语音情感识别的学习局部到全局特征聚合链接：https://arxiv.org/abs/2306.01491作者：Cheng Lu,Hailun Lian,Wenming Zheng,Yuan Zong,Yan Zhao,Sunan Li机构：Key Laboratory of Child Development and Learning Science of Ministry of Education, School of Biological Science and Medical Engineering, Southeast University, Nanjing, China, School of Information Science and Engineering, Southeast University, Nanjing, China备注：This paper has been accepted on INTERSPEECH 2023【4】 Towards Robust FastSpeech 2 by Modelling Residual Multimodality标题：用残差多模模型实现稳健的FastSpeech 2链接：https://arxiv.org/abs/2306.01442作者：Fabian Kögel,Bac Nguyen,Fabien Cardinaux机构：Sony Europe B.V., Stuttgart Laboratory , Germany备注：Accepted at INTERSPEECH 2023【5】 Improved DeepFake Detection Using Whisper Features标题：利用耳语特征改进的DeepFake检测链接：https://arxiv.org/abs/2306.01428作者：Piotr Kawa,Marcin Plata,Michał Czuba,Piotr Szymański,Piotr Syga机构：Wrocław University of Science and Technology, Wrocław, Poland备注：Accepted to INTERSPEECH 2023【6】 Speech Translation with Foundation Models and Optimal Transport: UPC at IWSLT23标题：基于基础模型和最优传输的语音翻译：UPC在IWSLT23上链接：https://arxiv.org/abs/2306.01327作者：Ioannis Tsiamas,Gerard I. Gállego,José A. R. Fonollosa,Marta R. Costa-jussà机构：Universitat Politècnica de Catalunya, Barcelona, Meta AI, Paris备注：IWSLT 2023【7】 JEPOO: Highly Accurate Joint Estimation of Pitch, Onset and Offset for Music Information Retrieval标题：JEPOO：用于音乐信息检索的高精度基音、首音和偏置联合估计链接：https://arxiv.org/abs/2306.01304作者：Haojie Wei,Jun Yuan,Rui Zhang,Yueguo Chen,Gang Wang机构：School of Information, Renmin University of China, Beijing, China, Huawei Noah’s Ark Lab, Shenzhen, China, Tsinghua University (www.ruizhang.info)备注：This paper has been accepted by IJCAI 2023; 11 pages, 6 figures【8】 DistilXLSR: A Light Weight Cross-Lingual Speech Representation Model标题：DistilXLSR：一种轻量级跨语言语音表示模型链接：https://arxiv.org/abs/2306.01303作者：Haoyu Wang,Siyuan Wang,Wei-Qiang Zhang,Jinfeng Bai机构：Department of Electronic Engineering, Tsinghua University, Beijing , China, TAL Education, Beijing , China备注：Accepted by INTERSPEECH 2023【9】 Learning When to Speak: Latency and Quality Trade-offs for Simultaneous Speech-to-Speech Translation with Offline Models标题：学习何时发言：同步语音到语音翻译与离线模型的延迟和质量权衡链接：https://arxiv.org/abs/2306.01201作者：Liam Dugan,Anshul Wadhawan,Kyle Spence,Chris Callison-Burch,Morgan McGuire,Victor Zordan机构：University of Pennsylvania,Roblox备注：To appear at INTERSPEECH 2023【10】 Exploration on HuBERT with Multiple Resolutions标题：关于休BERT的多分辨率问题探讨链接：https://arxiv.org/abs/2306.01084作者：Jiatong Shi,Yun Tang,Hirofumi Inaguma,Hongyu GOng,Juan Pino,Shinji Watanabe机构：Language Technologies Institute, Carnegie Mellon University, Meta AI备注：Accepted to Interspeech2023【11】 Bypass Temporal Classification: Weakly Supervised Automatic Speech Recognition with Imperfect Transcripts标题：绕过时间分类：弱监督语音识别链接：https://arxiv.org/abs/2306.01031作者：Dongji Gao,Matthew Wiesner,Hainan Xu,Leibny Paola Garcia,Daniel Povey,Sanjeev Khudanpur机构：CLSP & ,HLTCOE, Johns Hopkins University, USA, NVIDIA, USA, Xiaomi Corp., China【12】 How to Estimate Model Transferability of Pre-Trained Speech Models?标题：如何估计预先训练的语音模型的模型可转移性？链接：https://arxiv.org/abs/2306.01015作者：Zih-Ching Chen,Chao-Han Huck Yang,Bo Li,Yu Zhang,Nanxin Chen,Shou-Yiin Chang,Rohit Prabhavalkar,Hung-yi Lee,Tara N. Sainath机构：National Taiwan University, Taiwan, Google, USA, Georgia Tech, USA备注：Accepted to Interspeech. Code will be released【13】 Auditory Representation Effective for Estimating Vocal Tract Information标题：听觉表征在声道信息估计中的应用链接：https://arxiv.org/abs/2306.01522作者：Toshio Irino,Shintaro Doan机构：Wakayama University, Japan备注：This manuscript was submitted to APSIPA ASC 2023 on 2 Jun 2023【14】 Zero-Shot Blind Audio Bandwidth Extension标题：零触发盲音频带宽展宽链接：https://arxiv.org/abs/2306.01433作者：Eloi Moliner,Filip Elvander,Vesa Välimäki机构：V¨alim¨aki are with the Department ofInformation and Communications Engineering, Aalto University备注：Submitted to IEEE/ACM Transactions on Audio, Speech and Language Processing【15】 HD-DEMUCS: General Speech Restoration with Heterogeneous Decoders标题：HD-DEMUCS：基于异质解码器的通用语音恢复链接：https://arxiv.org/abs/2306.01411作者：Doyeon Kim,Soo-Whan Chung,Hyewon Han,Youna Ji,Hong-Goo Kang机构：Department of Electrical and Electronic Engineering, Yonsei University, South Korea, NAVER Cloud, South Korea备注：Accepted by INTERSPEECH 2023【16】 Task-Agnostic Structured Pruning of Speech Representation Models标题：与任务无关的语音表示模型结构化剪枝链接：https://arxiv.org/abs/2306.01385作者：Haoyu Wang,Siyuan Wang,Wei-Qiang Zhang,Hongbin Suo,Yulong Wan机构：Department of Electronic Engineering, Tsinghua University, Beijing , China, Data & AI Engineering System, OPPO, Beijing , China备注：Accepted by INTERSPEECH 2023【17】 Differentiable Grey-box Modelling of Phaser Effects using Frame-based Spectral Processing标题：基于帧的频谱处理相移效应的可微灰箱模型链接：https://arxiv.org/abs/2306.01332作者：Alistair Carson,Cassia Valentini-Botinhao,Simon King,Stefan Bilbao机构：Acoustics and Audio Group, University of Edinburgh, Edinburgh, UK, Centre for Speech Technology Research备注：Accepted for publication in Proc. DAFx23, Copenhagen, Denmark, September 2023【18】 Adapting an Unadaptable ASR System标题：适应不适应的ASR系统链接：https://arxiv.org/abs/2306.01208作者：Rao Ma,Mengjie Qian,Mark J. F. Gales,Kate M. Knill机构：ALTA Institute, Machine Intelligence Lab, Department of Engineering, Cambridge University, UK备注：submitted to INTERSPEECH【19】 ALO-VC: Any-to-any Low-latency One-shot Voice Conversion标题：Alo-VC：任意对任意低延迟单次语音转换链接：https://arxiv.org/abs/2306.01100作者：Bohan Wang,Damien Ronssin,Milos Cernak机构：´Ecole Polytechnique F´ed´erale de Lausanne, Lausanne, Switzerland, Logitech Europe S.A., Lausanne, Switzerland备注：Accepted to Interspeech 2023. Some audio samples are available at this https URL【20】 Adaptive ship-radiated noise recognition with learnable fine-grained wavelet transform标题：基于可学习细粒度小波变换的自适应舰船辐射噪声识别链接：https://arxiv.org/abs/2306.01002作者：Yuan Xie,Jiawei Ren,Ji Xu机构：Key laboratory of Speech Acoustics and Content Understanding, Institute of Acoustics, Chinese Academy of Sciences, University of Chinese Academy of Sciences备注：None【21】 Towards Selection of Text-to-speech Data to Augment ASR Training标题：选择文本到语音的数据以增强ASR训练链接：https://arxiv.org/abs/2306.00998作者：Shuo Liu,Leda Sarı,Chunyang Wu,Gil Keren,Yuan Shangguan,Jay Mahadeokar,Ozlem Kalinli机构：Meta AI, University of Augsburg, Augsburg, Germany【22】 Weakly-supervised forced alignment of disfluent speech using phoneme-level modeling标题：基于音素级建模的弱监督强对齐不流利语音链接：https://arxiv.org/abs/2306.00996作者：Theodoros Kouzelis,Georgios Paraskevopoulos,Athanasios Katsamanis,Vassilis Katsouros机构：Institute for Language and Speech Processing, Athena Research Center, School of ECE, National Technical University of Athens, Athens, Greece备注：Interspeech 2023eess.AS音频处理【1】 On Crowdsourcing-design with Comparison Category Rating for Evaluating Speech Enhancement Algorithms标题：评估语音增强算法的比较类别评分的众包设计链接：https://arxiv.org/abs/2306.01538作者：Angélica S. Z. Suárez,Clément Laroche,Line H. Clemmensen,Sneha Das机构：Dept. of Applied Mathematics and Computer Science, Technical University of Denmark, Denmark, GN Audio Research, Ballerup, Denmark备注：Published at ICASSP 2023【2】 Auditory Representation Effective for Estimating Vocal Tract Information标题：听觉表征在声道信息估计中的应用链接：https://arxiv.org/abs/2306.01522作者：Toshio Irino,Shintaro Doan机构：Wakayama University, Japan备注：This manuscript was submitted to APSIPA ASC 2023 on 2 Jun 2023【3】 Zero-Shot Blind Audio Bandwidth Extension标题：零触发盲音频带宽展宽链接：https://arxiv.org/abs/2306.01433作者：Eloi Moliner,Filip Elvander,Vesa Välimäki机构：V¨alim¨aki are with the Department ofInformation and Communications Engineering, Aalto University备注：Submitted to IEEE/ACM Transactions on Audio, Speech and Language Processing【4】 Audio-Visual Speech Enhancement with Score-Based Generative Models标题：基于分数生成模型的视听语音增强链接：https://arxiv.org/abs/2306.01432作者：Julius Richter,Simone Frintrop,Timo Gerkmann机构：Signal Processing Group,Computer Vision Group, Department of Informatics, Universität Hamburg, Germany备注：Submitted to ITG Conference on Speech Communication【5】 Active Noise Control in The New Century: The Role and Prospect of Signal Processing标题：新世纪的有源噪声控制：信号处理的作用和前景链接：https://arxiv.org/abs/2306.01425作者：Dongyuan Shi,Bhan Lam,Woon-Seng Gan,Jordan Cheer机构：School of Electrical & Electronic Engineering, Nanyang Technological University, Singapore, Institute of Sound and Vibration Research, University of Southampton, UK, Stepehen J. Elliott备注：Inter-Noise 2023【6】 HD-DEMUCS: General Speech Restoration with Heterogeneous Decoders标题：HD-DEMUCS：基于异质解码器的通用语音恢复链接：https://arxiv.org/abs/2306.01411作者：Doyeon Kim,Soo-Whan Chung,Hyewon Han,Youna Ji,Hong-Goo Kang机构：Department of Electrical and Electronic Engineering, Yonsei University, South Korea, NAVER Cloud, South Korea备注：Accepted by INTERSPEECH 2023【7】 Task-Agnostic Structured Pruning of Speech Representation Models标题：与任务无关的语音表示模型结构化剪枝链接：https://arxiv.org/abs/2306.01385作者：Haoyu Wang,Siyuan Wang,Wei-Qiang Zhang,Hongbin Suo,Yulong Wan机构：Department of Electronic Engineering, Tsinghua University, Beijing , China, Data & AI Engineering System, OPPO, Beijing , China备注：Accepted by INTERSPEECH 2023【8】 Differentiable Grey-box Modelling of Phaser Effects using Frame-based Spectral Processing标题：基于帧的频谱处理相移效应的可微灰箱模型链接：https://arxiv.org/abs/2306.01332作者：Alistair Carson,Cassia Valentini-Botinhao,Simon King,Stefan Bilbao机构：Acoustics and Audio Group, University of Edinburgh, Edinburgh, UK, Centre for Speech Technology Research备注：Accepted for publication in Proc. DAFx23, Copenhagen, Denmark, September 2023【9】 Improved Training for End-to-End Streaming Automatic Speech Recognition Model with Punctuation标题：改进的带标点符号的端到端流自动语音识别模型训练链接：https://arxiv.org/abs/2306.01296作者：Hanbyul Kim,Seunghyun Seo,Lukas Lee,Seolki Baek机构：NAVER Cloud, South Korea备注：Accepted at INTERSPEECH 2023【10】 Tensor decomposition for minimization of E2E SLU model toward on-device processing标题：用于最小化E2E SLU模型的张量分解面向在设备处理链接：https://arxiv.org/abs/2306.01247作者：Yosuke Kashiwagi,Siddhant Arora,Hayato Futami,Jessica Huynh,Shih-Lun Wu,Yifan Peng,Brian Yan,Emiru Tsunoo,Shinji Watanabe机构：Sony Group Corporation, Japan, Carnegie Mellon University, USA备注：Accepted by INTERSPEECH 2023【11】 Adapting an Unadaptable ASR System标题：适应不适应的ASR系统链接：https://arxiv.org/abs/2306.01208作者：Rao Ma,Mengjie Qian,Mark J. F. Gales,Kate M. Knill机构：ALTA Institute, Machine Intelligence Lab, Department of Engineering, Cambridge University, UK备注：submitted to INTERSPEECH【12】 ALO-VC: Any-to-any Low-latency One-shot Voice Conversion标题：Alo-VC：任意对任意低延迟单次语音转换链接：https://arxiv.org/abs/2306.01100作者：Bohan Wang,Damien Ronssin,Milos Cernak机构：´Ecole Polytechnique F´ed´erale de Lausanne, Lausanne, Switzerland, Logitech Europe S.A., Lausanne, Switzerland备注：Accepted to Interspeech 2023. Some audio samples are available at this https URL【13】 Adaptive ship-radiated noise recognition with learnable fine-grained wavelet transform标题：基于可学习细粒度小波变换的自适应舰船辐射噪声识别链接：https://arxiv.org/abs/2306.01002作者：Yuan Xie,Jiawei Ren,Ji Xu机构：Key laboratory of Speech Acoustics and Content Understanding, Institute of Acoustics, Chinese Academy of Sciences, University of Chinese Academy of Sciences备注：None【14】 Towards Selection of Text-to-speech Data to Augment ASR Training标题：选择文本到语音的数据以增强ASR训练链接：https://arxiv.org/abs/2306.00998作者：Shuo Liu,Leda Sarı,Chunyang Wu,Gil Keren,Yuan Shangguan,Jay Mahadeokar,Ozlem Kalinli机构：Meta AI, University of Augsburg, Augsburg, Germany【15】 Weakly-supervised forced alignment of disfluent speech using phoneme-level modeling标题：基于音素级建模的弱监督强对齐不流利语音链接：https://arxiv.org/abs/2306.00996作者：Theodoros Kouzelis,Georgios Paraskevopoulos,Athanasios Katsamanis,Vassilis Katsouros机构：Institute for Language and Speech Processing, Athena Research Center, School of ECE, National Technical University of Athens, Athens, Greece备注：Interspeech 2023【16】 Q&A: Query-Based Representation Learning for Multi-Track Symbolic Music re-Arrangement标题：问答：基于查询的多声道符号音乐重排表示学习链接：https://arxiv.org/abs/2306.01635作者：Jingwei Zhao,Gus Xia,Ye Wang机构：School of Computing, NUS, Institute of Data Science, NUS, Integrative Sciences and Engineering Programme, NUS Graduate School, Music X Lab, NYU Shanghai, MBZUAI备注：Accepted by IJCAI 2023 Special Track for AI the Arts and Creativity【17】 Enhance Temporal Relations in Audio Captioning with Sound Event Detection标题：利用声音事件检测增强音频字幕中的时间关系链接：https://arxiv.org/abs/2306.01533作者：Zeyu Xie,Xuenan Xu,Mengyue Wu,Kai Yu机构：MoE Key Lab of Artificial Intelligence X-LANCE Lab, Department of Computer Science and Engineering AI Institute, Shanghai Jiao Tong University, Shanghai, China【18】 BabySLM: language-acquisition-friendly benchmark of self-supervised spoken language models标题：BabySLM：自我监督口语模型的语言习得友好基准链接：https://arxiv.org/abs/2306.01506作者：Marvin Lavechin,Yaya Sy,Hadrien Titeux,María Andrea Cruz Blandón,Okko Räsänen,Hervé Bredin,Emmanuel Dupoux,Alejandrina Cristia机构：LSCP, ENS, EHESS, CNRS, PSL University, Paris, France, Meta AI Research, France, Unit of Computing Sciences, Tampere University, Finland, IRIT, CNRS, Toulouse, France, Cognitive Machine Learning Team, INRIA, France备注：Proceedings of Interspeech 2023【19】 Towards Robust FastSpeech 2 by Modelling Residual Multimodality标题：用残差多模模型实现稳健的FastSpeech 2链接：https://arxiv.org/abs/2306.01442作者：Fabian Kögel,Bac Nguyen,Fabien Cardinaux机构：Sony Europe B.V., Stuttgart Laboratory , Germany备注：Accepted at INTERSPEECH 2023【20】 Improved DeepFake Detection Using Whisper Features标题：利用耳语特征改进的DeepFake检测链接：https://arxiv.org/abs/2306.01428作者：Piotr Kawa,Marcin Plata,Michał Czuba,Piotr Szymański,Piotr Syga机构：Wrocław University of Science and Technology, Wrocław, Poland备注：Accepted to INTERSPEECH 2023【21】 Speech Translation with Foundation Models and Optimal Transport: UPC at IWSLT23标题：基于基础模型和最优传输的语音翻译：UPC在IWSLT23上链接：https://arxiv.org/abs/2306.01327作者：Ioannis Tsiamas,Gerard I. Gállego,José A. R. Fonollosa,Marta R. Costa-jussà机构：Universitat Politècnica de Catalunya, Barcelona, Meta AI, Paris备注：IWSLT 2023【22】 JEPOO: Highly Accurate Joint Estimation of Pitch, Onset and Offset for Music Information Retrieval标题：JEPOO：用于音乐信息检索的高精度基音、首音和偏置联合估计链接：https://arxiv.org/abs/2306.01304作者：Haojie Wei,Jun Yuan,Rui Zhang,Yueguo Chen,Gang Wang机构：School of Information, Renmin University of China, Beijing, China, Huawei Noah’s Ark Lab, Shenzhen, China, Tsinghua University (www.ruizhang.info)备注：This paper has been accepted by IJCAI 2023; 11 pages, 6 figures【23】 DistilXLSR: A Light Weight Cross-Lingual Speech Representation Model标题：DistilXLSR：一种轻量级跨语言语音表示模型链接：https://arxiv.org/abs/2306.01303作者：Haoyu Wang,Siyuan Wang,Wei-Qiang Zhang,Jinfeng Bai机构：Department of Electronic Engineering, Tsinghua University, Beijing , China, TAL Education, Beijing , China备注：Accepted by INTERSPEECH 2023【24】 Learning When to Speak: Latency and Quality Trade-offs for Simultaneous Speech-to-Speech Translation with Offline Models标题：学习何时发言：同步语音到语音翻译与离线模型的延迟和质量权衡链接：https://arxiv.org/abs/2306.01201作者：Liam Dugan,Anshul Wadhawan,Kyle Spence,Chris Callison-Burch,Morgan McGuire,Victor Zordan机构：University of Pennsylvania,Roblox备注：To appear at INTERSPEECH 2023【25】 Exploration on HuBERT with Multiple Resolutions标题：关于休BERT的多分辨率问题探讨链接：https://arxiv.org/abs/2306.01084作者：Jiatong Shi,Yun Tang,Hirofumi Inaguma,Hongyu GOng,Juan Pino,Shinji Watanabe机构：Language Technologies Institute, Carnegie Mellon University, Meta AI备注：Accepted to Interspeech2023【26】 Bypass Temporal Classification: Weakly Supervised Automatic Speech Recognition with Imperfect Transcripts标题：绕过时间分类：弱监督语音识别链接：https://arxiv.org/abs/2306.01031作者：Dongji Gao,Matthew Wiesner,Hainan Xu,Leibny Paola Garcia,Daniel Povey,Sanjeev Khudanpur机构：CLSP & ,HLTCOE, Johns Hopkins University, USA, NVIDIA, USA, Xiaomi Corp., China【27】 How to Estimate Model Transferability of Pre-Trained Speech Models?标题：如何估计预先训练的语音模型的模型可转移性？链接：https://arxiv.org/abs/2306.01015作者：Zih-Ching Chen,Chao-Han Huck Yang,Bo Li,Yu Zhang,Nanxin Chen,Shou-Yiin Chang,Rohit Prabhavalkar,Hung-yi Lee,Tara N. Sainath机构：National Taiwan University, Taiwan, Google, USA, Georgia Tech, USA备注：Accepted to Interspeech. Code will be released机器翻译，仅供参考
每日语音速递
