2021 IEEE Spoken Language Technology Workshop (SLT)最新文献_第10页

How Far Are We from Robust Voice Conversion: A Survey 我们离稳健的语音转换还有多远

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2020-11-24 DOI: 10.1109/SLT48900.2021.9383498

Tzu-hsien Huang, Jheng-hao Lin, Chien-yu Huang, Hung-yi Lee

引用次数: 15

Two-Way Neural Machine Translation: A Proof of Concept for Bidirectional Translation Modeling Using a Two-Dimensional Grid 双向神经机器翻译：使用二维网格进行双向翻译建模的概念验证

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2020-11-24 DOI: 10.1109/SLT48900.2021.9383589

Parnia Bahar, Christopher Brix, H. Ney

引用次数: 1

End-To-End Silent Speech Recognition with Acoustic Sensing 端到端静音语音识别与声学传感

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2020-11-23 DOI: 10.1109/SLT48900.2021.9383622

Jian Luo, Jianzong Wang, Ning Cheng, Guilin Jiang, Jing Xiao

引用次数: 3

Self-Supervised Learning with Cross-Modal Transformers for Emotion Recognition 基于跨模态变换的自监督学习用于情绪识别

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2020-11-20 DOI: 10.1109/SLT48900.2021.9383618

Aparna Khare, Srinivas Parthasarathy, Shiva Sundaram

引用次数: 25

Tal: A Synchronised Multi-Speaker Corpus of Ultrasound Tongue Imaging, Audio, and Lip Videos 一个同步的多扬声器语料库，包括舌头超声成像、音频和嘴唇视频

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2020-11-19 DOI: 10.1109/SLT48900.2021.9383619

M. Ribeiro, Jennifer Sanger, Jingxuan Zhang, Aciel Eshky, A. Wrench, Korin Richmond, S. Renals

引用次数: 21

Multi-Channel Automatic Speech Recognition Using Deep Complex Unet 基于深度复杂网络的多通道自动语音识别

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2020-11-18 DOI: 10.1109/SLT48900.2021.9383492

Yuxiang Kong, Jian Wu, Quandong Wang, Peng Gao, Weiji Zhuang, Yujun Wang, Lei Xie

{"title":"Multi-Channel Automatic Speech Recognition Using Deep Complex Unet","authors":"Yuxiang Kong, Jian Wu, Quandong Wang, Peng Gao, Weiji Zhuang, Yujun Wang, Lei Xie","doi":"10.1109/SLT48900.2021.9383492","DOIUrl":"https://doi.org/10.1109/SLT48900.2021.9383492","url":null,"abstract":"The front-end module in multi-channel automatic speech recognition (ASR) systems mainly use microphone array techniques to produce enhanced signals in noisy conditions with reverberation and echos. Recently, neural network (NN) based front-end has shown promising improvement over the conventional signal processing methods. In this paper, we propose to adopt the architecture of deep complex Unet (DCUnet) - a powerful complex-valued Unet-structured speech enhancement model - as the front-end of the multi-channel acoustic model, and integrate them in a multi-task learning (MTL) framework along with cascaded framework for comparison. Meanwhile, we investigate the proposed methods with several training strategies to improve the recognition accuracy on the 1000-hours real-world XiaoMi smart speaker data with echos. Experiments show that our proposed DCUnet-MTL method brings about 12.2% relative character error rate (CER) reduction compared with the traditional approach with array processing plus single-channel acoustic model. It also achieves superior performance than the recently proposed neural beamforming method.","PeriodicalId":243211,"journal":{"name":"2021 IEEE Spoken Language Technology Workshop (SLT)","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114211936","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 4

On the Use of Self-Supervised Pre-Trained Acoustic and Linguistic Features for Continuous Speech Emotion Recognition 自监督预训练声学和语言特征在连续语音情感识别中的应用

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2020-11-18 DOI: 10.1109/SLT48900.2021.9383456

Manon Macary, Marie Tahon, Y. Estève, Anthony Rousseau

引用次数: 32

WPD++: An Improved Neural Beamformer for Simultaneous Speech Separation and Dereverberation 一种用于同时语音分离和去噪的改进神经波束形成器

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2020-11-18 DOI: 10.1109/SLT48900.2021.9383528

Zhaoheng Ni, Yong Xu, Meng Yu, Bo Wu, Shi-Xiong Zhang, Dong Yu, Michael I. Mandel

{"title":"WPD++: An Improved Neural Beamformer for Simultaneous Speech Separation and Dereverberation","authors":"Zhaoheng Ni, Yong Xu, Meng Yu, Bo Wu, Shi-Xiong Zhang, Dong Yu, Michael I. Mandel","doi":"10.1109/SLT48900.2021.9383528","DOIUrl":"https://doi.org/10.1109/SLT48900.2021.9383528","url":null,"abstract":"This paper aims at eliminating the interfering speakers' speech, additive noise, and reverberation from the noisy multi-talker speech mixture that benefits automatic speech recognition (ASR) backend. While the recently proposed Weighted Power minimization Distortionless response (WPD) beamformer can perform separation and dereverberation simultaneously, the noise cancellation component still has the potential to progress. We propose an improved neural WPD beamformer called \"WPD++\" by an enhanced beamforming module in the conventional WPD and a multi-objective loss function for the joint training. The beamforming module is improved by utilizing the spatio-temporal correlation. A multi-objective loss, including the complex spectra domain scale-invariant signal-to-noise ratio (C-Si-SNR) and the magnitude domain mean square error (Mag-MSE), is properly designed to make multiple constraints on the enhanced speech and the desired power of the dry clean signal. Joint training is conducted to optimize the complex-valued mask estimator and the WPD++ beamformer in an end-to-end way. The results show that the proposed WPD++ outperforms several state-of-the-art beamformers on the enhanced speech quality and word error rate (WER) of ASR.","PeriodicalId":243211,"journal":{"name":"2021 IEEE Spoken Language Technology Workshop (SLT)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-11-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127053543","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 6

Learn2Sing: Target Speaker Singing Voice Synthesis by Learning from a Singing Teacher Learn2Sing:目标说话者从歌唱老师那里学习歌唱声音合成

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2020-11-17 DOI: 10.1109/SLT48900.2021.9383585

Heyang Xue, Shan Yang, Yinjiao Lei, Lei Xie, Xiulin Li

{"title":"Learn2Sing: Target Speaker Singing Voice Synthesis by Learning from a Singing Teacher","authors":"Heyang Xue, Shan Yang, Yinjiao Lei, Lei Xie, Xiulin Li","doi":"10.1109/SLT48900.2021.9383585","DOIUrl":"https://doi.org/10.1109/SLT48900.2021.9383585","url":null,"abstract":"Singing voice synthesis has been paid rising attention with the rapid development of speech synthesis area. In general, a studio-level singing corpus is usually necessary to produce a natural singing voice from lyrics and music-related transcription. However, such a corpus is difficult to collect since it’s hard for many of us to sing like a professional singer. In this paper, we propose an approach – Learn2Sing that only needs a singing teacher to generate the target speakers’ singing voice without their singing voice data. In our approach, a teacher’s singing corpus and speech from multiple target speakers are trained in a frame-level auto-regressive acoustic model where singing and speaking share the common speaker embedding and style tag embedding. Meanwhile, since there is no music-related transcription for the target speaker, we use log-scale fundamental frequency (LF0) as an auxiliary feature as the inputs of the acoustic model for building a unified input representation. In order to enable the target speaker to sing without singing reference audio in the inference stage, a duration model and an LF0 prediction model are also trained. Particularly, we employ domain adversarial training (DAT) in the acoustic model, which aims to enhance the singing performance of target speakers by disentangling style from acoustic features of singing and speaking data. Our experiments indicate that the proposed approach is capable of synthesizing singing voice for target speaker given only their speech samples.","PeriodicalId":243211,"journal":{"name":"2021 IEEE Spoken Language Technology Workshop (SLT)","volume":"670 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122968199","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 7

Cascade RNN-Transducer: Syllable Based Streaming On-Device Mandarin Speech Recognition with a Syllable-To-Character Converter 级联rnn -换能器:基于音节流的设备上普通话语音识别与音节-字符转换器

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2020-11-17 DOI: 10.1109/SLT48900.2021.9383506

Xiong Wang, Zhuoyuan Yao, Xian Shi, Lei Xie

{"title":"Cascade RNN-Transducer: Syllable Based Streaming On-Device Mandarin Speech Recognition with a Syllable-To-Character Converter","authors":"Xiong Wang, Zhuoyuan Yao, Xian Shi, Lei Xie","doi":"10.1109/SLT48900.2021.9383506","DOIUrl":"https://doi.org/10.1109/SLT48900.2021.9383506","url":null,"abstract":"End-to-end models are favored in automatic speech recognition (ASR) because of its simplified system structure and superior performance. Among these models, recurrent neural network transducer (RNN-T) has achieved significant progress in streaming on-device speech recognition because of its high-accuracy and low-latency. RNN-T adopts a prediction network to enhance language information, but its language modeling ability is limited because it still needs paired speech-text data to train. Further strengthening the language modeling ability through extra text data, such as shallow fusion with an external language model, only brings a small performance gain. In view of the fact that Mandarin Chinese is a character-based language and each character is pronounced as a tonal syllable, this paper proposes a novel cascade RNN-T approach to improve the language modeling ability of RNN-T. Our approach firstly uses an RNN-T to transform acoustic feature into syllable sequence, and then converts the syllable sequence into character sequence through an RNN-T-based syllable-to-character converter. Thus a rich text repository can be easily used to strengthen the language model ability. By introducing several important tricks, the cascade RNN-T approach surpasses the character-based RNN-T by a large margin on several Mandarin test sets, with much higher recognition quality and similar latency.","PeriodicalId":243211,"journal":{"name":"2021 IEEE Spoken Language Technology Workshop (SLT)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122413104","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 25