语音/音频处理学术速递[8.8]

今日论文合集：cs.SD语音17篇，eess.AS音频处理17篇。

cs.SD语音

【1】 Improving Deep Attractor Network by BGRU and GMM for Speech Separation标题：用BGRU和GMM改进深度吸引子网络进行语音分离链接：https://arxiv.org/abs/2308.03332

作者：Rawad Melhem,Assef Jafar,Riad Hamadeh备注：None

【2】 Do You Remember? Overcoming Catastrophic Forgetting for Fake Audio Detection标题：你还记得吗？克服伪音频检测的灾难性遗忘链接：https://arxiv.org/abs/2308.03300

作者：Xiaohui Zhang,Jiangyan Yi,Jianhua Tao,Chenglong Wang,Chuyuan Zhang机构： 1State Key Laboratory of Multimodal Artificial IntelligenceSystems, Institute of Automation, China 2School of Computer and Information Technology, University of Beijing Jiaotong, China 3Department ofAutomation, Tsinghua University备注：40th Internation Conference on Machine Learning (ICML 2023)

【3】 SeACo-Paraformer: A Non-Autoregressive ASR System with Flexible and Effective Hotword Customization Ability标题：SeaCo-Paraform：一种具有灵活有效的关键词定制能力的非自回归ASR系统链接：https://arxiv.org/abs/2308.03266

作者：Xian Shi,Yexin Yang,Zerui Li,Shiliang Zhang机构：Speech Lab of DAMO Academy, Alibaba Group, China备注：early draft

【4】 Characterization of cough sounds using statistical analysis标题：利用统计分析对咳嗽音进行表征链接：https://arxiv.org/abs/2308.03019

作者：Naveenkumar Vodnala,Pratap Reddy Lankireddy,Padmasai Yarlagadda机构：Department of Electronics and Communication Engineering; bVNR Vignana Jyothi Institute of, Engineering and Technology; cJawaharlal Nehru Technological University; dHyderabad备注：19 pages, 8 figures, paper submitted to journal Biomedical Signal Processing and Control which is under review

【5】 DiffDance: Cascaded Human Motion Diffusion Model for Dance Generation标题：DiffDance：面向舞蹈生成的级联人体运动扩散模型链接：https://arxiv.org/abs/2308.02915

作者：Qiaosong Qi,Le Zhuo,Aixi Zhang,Yue Liao,Fei Fang,Si Liu,Shuicheng Yan机构：Alibaba Group, Beijing, China, Beihang University, BAAI, Skyworks, Ballet Jazz, House, Krump, Waack备注：Accepted at ACM MM 2023

【6】 Elucidate Gender Fairness in Singing Voice Transcription标题：歌声改编中的性别公平问题链接：https://arxiv.org/abs/2308.02898

作者：Xiangming Gu,Wei Zeng,Ye Wang机构：Integrative Sciences and Engineering, Programme, NUS Graduate School, National University of Singapore, School of Computing, National备注：Camera-ready version of ACM MM2023

【7】 ApproBiVT: Lead ASR Models to Generalize Better Using Approximated Bias-Variance Tradeoff Guided Early Stopping and Checkpoint Averaging标题：ApproBiVT：使用近似偏差-方差权衡指导提前停止和检查点平均，引导ASR模型更好地泛化链接：https://arxiv.org/abs/2308.02870

作者：Fangyuan Wang,Ming Hao,Yuhai Shi,Bo Xu机构： Institute of Automation, Chinese Academy of Sciences, Beijing, China, Academy of Broadcasting Science, National Radio and Television Administration, Beijing, School of Future Technology, University of Chinese Academy of Sciences , Beijing, China

【8】 A Systematic Exploration of Joint-training for Singing Voice Synthesis标题：歌唱嗓音合成联合训练的系统探索链接：https://arxiv.org/abs/2308.02867

作者：Yuning Wu,Yifeng Yu,Jiatong Shi,Tao Qian,Qin Jin机构： Renmin University of China, Georgia Institute of Technology, Carnegie Mellon University

【9】 Bootstrapping Contrastive Learning Enhanced Music Cold-Start Matching标题：自举对比学习增强型音乐冷启动匹配链接：https://arxiv.org/abs/2308.02844

作者：Xinping Zhao,Ying Zhang,Qiang Xiao,Yuming Ren,Yingchun Yang机构：Zhejiang University, NetEase Cloud Music, NetEase Inc., Hangzhou, China备注：None

【10】 Towards Improving Harmonic Sensitivity and Prediction Stability for Singing Melody Extraction标题：提高歌唱旋律提取的谐波灵敏度和预测稳定性的研究链接：https://arxiv.org/abs/2308.02723

作者：Keren Shao,Ke Chen,Taylor Berg-Kirkpatrick,Shlomo Dubnov机构：University of California San Diego备注：7 pages, 4 figures, 2 tables, Proceedings of the 24th International Society for Music Information Retrieval Conference, ISMIR 2023

【11】 From Discrete Tokens to High-Fidelity Audio Using Multi-Band Diffusion标题：利用多带扩散从离散令牌到高保真音频链接：https://arxiv.org/abs/2308.02560

作者：Robin San Roman,Yossi Adi,Antoine Deleforge,Romain Serizel,Gabriel Synnaeve,Alexandre Défossez机构：Alexandre Defossez♢, ♢ : Meta AI, ♠: Universite de Lorraine, CNRS, Inria, LORIA, Nancy, France, ♣ : Hebrew University of Jerusalem.备注：10 pages

【12】 Active Noise Control based on the Momentum Multichannel Normalized Filtered-x Least Mean Square Algorithm标题：基于动量多通道归一化滤波x最小均方算法的有源噪声控制链接：https://arxiv.org/abs/2308.03684

作者：Dongyuan Shi,Woon-Seng Gan,Bhan Lam,Shulin Wen,Xiaoyi Shen机构：School of Electrical and Electronic Engineering, Nanyang Technological University., Nanyang Ave, Singapore ,.备注：Conference: INTER-NOISE and NOISE-CON Congress and Conference Proceedings 2020 At Korea Volume: 261

【13】 AudioVMAF: Audio Quality Prediction with VMAF标题：AudioVMAF：使用VMAF预测音频质量链接：https://arxiv.org/abs/2308.03437

作者：Arijit Biswas,Harald Mundt机构：Sound Tech Research, Dolby Germany GmbH备注：Accepted to 155th Audio Engineering Society (AES), New York, NY, USA, October 2023

【14】 Investigation of Self-supervised Pre-trained Models for Classification of Voice Quality from Speech and Neck Surface Accelerometer Signals标题：用于语音和颈部加速度计信号语音质量分类的自监督预训练模型研究链接：https://arxiv.org/abs/2308.03226

作者：Sudarsana Reddy Kadiri,Farhad Javanmardi,Paavo Alku机构：Department of Information and Communications Engineering, Aalto University, Finland备注：Accepted by Computer Speech & Language

【15】 Ensemble Distillation Network: Learning Robust Speaker Representations without Supervision标题：集成蒸馏网络：无监督学习稳健说话人表示链接：https://arxiv.org/abs/2308.02774

作者：Yafeng Chen,Siqi Zheng,Qian Chen机构：Speech Lab of DAMO Academy, Alibaba Group备注：arXiv admin note: text overlap with arXiv:2211.04168

【16】 Choir Transformer: Generating Polyphonic Music with Relative Attention on Transformer标题：《合唱Transformer》：在《Transformer》上相对关注地产生复调音乐链接：https://arxiv.org/abs/2308.02531

作者：Jiuyang Zhou,Hong Zhu,Xingping Wang机构：School of Automation and Information Engineering, Xi’an University of Technology

【17】 MyVoice: Arabic Speech Resource Collaboration Platform标题：MyVoice：阿拉伯语语音资源协作平台链接：https://arxiv.org/abs/2308.02503

作者：Yousseif Elshahawy,Yassine El Kheir,Shammur Absar Chowdhury,Ahmed Ali机构：Qatar Computing Research Institute, HBKU, Doha, Qatar备注：2 pages, accepted at InterSpeech23 Show and Tell Session

eess.AS音频处理

【1】 Active Noise Control based on the Momentum Multichannel Normalized Filtered-x Least Mean Square Algorithm标题：基于动量多通道归一化滤波x最小均方算法的有源噪声控制链接：https://arxiv.org/abs/2308.03684

【2】 AudioVMAF: Audio Quality Prediction with VMAF标题：AudioVMAF：使用VMAF预测音频质量链接：https://arxiv.org/abs/2308.03437

作者：Arijit Biswas,Harald Mundt机构：Sound Tech Research, Dolby Germany GmbH备注：Accepted to 155th Audio Engineering Society (AES), New York, NY, USA, October 2023

【3】 Investigation of Self-supervised Pre-trained Models for Classification of Voice Quality from Speech and Neck Surface Accelerometer Signals标题：用于语音和颈部加速度计信号语音质量分类的自监督预训练模型研究链接：https://arxiv.org/abs/2308.03226

作者：Sudarsana Reddy Kadiri,Farhad Javanmardi,Paavo Alku机构：Department of Information and Communications Engineering, Aalto University, Finland备注：Accepted by Computer Speech & Language

【4】 Ensemble Distillation Network: Learning Robust Speaker Representations without Supervision标题：集成蒸馏网络：无监督学习稳健说话人表示链接：https://arxiv.org/abs/2308.02774

作者：Yafeng Chen,Siqi Zheng,Qian Chen机构：Speech Lab of DAMO Academy, Alibaba Group备注：arXiv admin note: text overlap with arXiv:2211.04168

【5】 Choir Transformer: Generating Polyphonic Music with Relative Attention on Transformer标题：《合唱Transformer》：在《Transformer》上相对关注地产生复调音乐链接：https://arxiv.org/abs/2308.02531

作者：Jiuyang Zhou,Hong Zhu,Xingping Wang机构：School of Automation and Information Engineering, Xi’an University of Technology

【6】 MyVoice: Arabic Speech Resource Collaboration Platform标题：MyVoice：阿拉伯语语音资源协作平台链接：https://arxiv.org/abs/2308.02503

【7】 Improving Deep Attractor Network by BGRU and GMM for Speech Separation标题：用BGRU和GMM改进深度吸引子网络进行语音分离链接：https://arxiv.org/abs/2308.03332

作者：Rawad Melhem,Assef Jafar,Riad Hamadeh备注：None

【8】 Do You Remember? Overcoming Catastrophic Forgetting for Fake Audio Detection标题：你还记得吗？克服伪音频检测的灾难性遗忘链接：https://arxiv.org/abs/2308.03300

【9】 SeACo-Paraformer: A Non-Autoregressive ASR System with Flexible and Effective Hotword Customization Ability标题：SeaCo-Paraform：一种具有灵活有效的关键词定制能力的非自回归ASR系统链接：https://arxiv.org/abs/2308.03266

作者：Xian Shi,Yexin Yang,Zerui Li,Shiliang Zhang机构：Speech Lab of DAMO Academy, Alibaba Group, China备注：early draft

【10】 Characterization of cough sounds using statistical analysis标题：利用统计分析对咳嗽音进行表征链接：https://arxiv.org/abs/2308.03019

【11】 DiffDance: Cascaded Human Motion Diffusion Model for Dance Generation标题：DiffDance：面向舞蹈生成的级联人体运动扩散模型链接：https://arxiv.org/abs/2308.02915

【12】 Elucidate Gender Fairness in Singing Voice Transcription标题：歌声改编中的性别公平问题链接：https://arxiv.org/abs/2308.02898

【13】 ApproBiVT: Lead ASR Models to Generalize Better Using Approximated Bias-Variance Tradeoff Guided Early Stopping and Checkpoint Averaging标题：ApproBiVT：使用近似偏差-方差权衡指导提前停止和检查点平均，引导ASR模型更好地泛化链接：https://arxiv.org/abs/2308.02870

【14】 A Systematic Exploration of Joint-training for Singing Voice Synthesis标题：歌唱嗓音合成联合训练的系统探索链接：https://arxiv.org/abs/2308.02867

作者：Yuning Wu,Yifeng Yu,Jiatong Shi,Tao Qian,Qin Jin机构： Renmin University of China, Georgia Institute of Technology, Carnegie Mellon University

【15】 Bootstrapping Contrastive Learning Enhanced Music Cold-Start Matching标题：自举对比学习增强型音乐冷启动匹配链接：https://arxiv.org/abs/2308.02844

作者：Xinping Zhao,Ying Zhang,Qiang Xiao,Yuming Ren,Yingchun Yang机构：Zhejiang University, NetEase Cloud Music, NetEase Inc., Hangzhou, China备注：None

【16】 Towards Improving Harmonic Sensitivity and Prediction Stability for Singing Melody Extraction标题：提高歌唱旋律提取的谐波灵敏度和预测稳定性的研究链接：https://arxiv.org/abs/2308.02723

【17】 From Discrete Tokens to High-Fidelity Audio Using Multi-Band Diffusion标题：利用多带扩散从离散令牌到高保真音频链接：https://arxiv.org/abs/2308.02560

机器翻译，仅供参考

语音/音频处理学术速递[8.8]

评论 0