2021 IEEE Spoken Language Technology Workshop (SLT)最新文献_第3页

Controllable Emphatic Speech Synthesis based on Forward Attention for Expressive Speech Synthesis 表达性语音合成中基于前向注意的可控重点语音合成

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2021-01-19 DOI: 10.1109/SLT48900.2021.9383537

Liangqi Liu, Jiankun Hu, Zhiyong Wu, Song Yang, Songfan Yang, Jia Jia, H. Meng

{"title":"Controllable Emphatic Speech Synthesis based on Forward Attention for Expressive Speech Synthesis","authors":"Liangqi Liu, Jiankun Hu, Zhiyong Wu, Song Yang, Songfan Yang, Jia Jia, H. Meng","doi":"10.1109/SLT48900.2021.9383537","DOIUrl":"https://doi.org/10.1109/SLT48900.2021.9383537","url":null,"abstract":"In speech interaction scenarios, speech emphasis is essential for expressing the underlying intention and attitude. Recently, end-to-end emphatic speech synthesis greatly improves the naturalness of synthetic speech, but also brings new problems: 1) lack of interpretability for how emphatic codes affect the model; 2) no separate control of emphasis on duration and on intonation and energy. We propose a novel way to build an interpretable and controllable emphatic speech synthesis framework based on forward attention. Firstly, we explicitly model the local variation of speaking rate for emphasized words and neutral words with modified forward attention to manifest emphasized words in terms of duration. The 2-layers LSTM in decoder is further divided into attention-RNN and decoder-RNN to disentangle the influence of emphasis on duration and on intonation and energy. The emphasis information is injected into decoder-RNN for highlighting emphasized words in the aspects of intonation and energy. Experimental results have shown that our model can not only provide separate control of emphasis on duration and on intonation and energy, but also generate more robust and prominent emphatic speech with high quality and naturalness.","PeriodicalId":243211,"journal":{"name":"2021 IEEE Spoken Language Technology Workshop (SLT)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129034703","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

Dynamically Weighted Ensemble Models for Automatic Speech Recognition 自动语音识别的动态加权集成模型

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2021-01-19 DOI: 10.1109/SLT48900.2021.9383463

K. Praveen, Abhishek Pandey, D. Kumar, S. Rath, Sandip Shriram Bapat

{"title":"Dynamically Weighted Ensemble Models for Automatic Speech Recognition","authors":"K. Praveen, Abhishek Pandey, D. Kumar, S. Rath, Sandip Shriram Bapat","doi":"10.1109/SLT48900.2021.9383463","DOIUrl":"https://doi.org/10.1109/SLT48900.2021.9383463","url":null,"abstract":"In machine learning, training multiple models for the same task, and using the outputs from all the models helps reduce the variance of the combined result. Using an ensemble of models in classification tasks such as Automatic Speech Recognition (ASR) improves the accuracy across different target domains such as multiple accents, environmental conditions, and other scenarios. It is possible to select model weights for the ensemble in numerous ways. A classifier trained to identify target domain, a simple averaging function, or an exhaustive grid search are the common approaches to obtain suitable weights. All these methods suffer either in choosing sub-optimal weights or by being computationally expensive. We propose a novel and practical method for dynamic weight selection in an ensemble, which can approximate a grid search in a time-efficient manner. We show that a combination of weights always performs better than assigning uniform weights for all models. Our algorithm can utilize a validation set if available or find weights dynamically from the input utterance itself. Experiments conducted for various ASR tasks show that the proposed method outperforms the uniformly weighted ensemble in terms of Word Error Rate (WER) in our experiments.","PeriodicalId":243211,"journal":{"name":"2021 IEEE Spoken Language Technology Workshop (SLT)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133612322","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: 1

Lightspeech: Lightweight Non-Autoregressive Multi-Speaker Text-To-Speech 轻语音:轻量级非自回归多说话文本到语音

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2021-01-19 DOI: 10.1109/SLT48900.2021.9383562

Song Li, Beibei Ouyang, Lin Li, Q. Hong

引用次数: 4

Towards Unsupervised Learning of Speech Features in the Wild 面向野外语音特征的无监督学习

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2021-01-19 DOI: 10.1109/SLT48900.2021.9383461

M. Rivière, Emmanuel Dupoux

引用次数: 20

Data-Filtering Methods for Self-Training of Automatic Speech Recognition Systems 自动语音识别系统自训练的数据滤波方法

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2021-01-19 DOI: 10.1109/SLT48900.2021.9383577

Alexandru-Lucian Georgescu, Cristian Manolache, Dan Oneaţă, H. Cucu, C. Burileanu

{"title":"Data-Filtering Methods for Self-Training of Automatic Speech Recognition Systems","authors":"Alexandru-Lucian Georgescu, Cristian Manolache, Dan Oneaţă, H. Cucu, C. Burileanu","doi":"10.1109/SLT48900.2021.9383577","DOIUrl":"https://doi.org/10.1109/SLT48900.2021.9383577","url":null,"abstract":"Self-training is a simple and efficient way of leveraging un-labeled speech data: (i) start with a seed system trained on transcribed speech; (ii) pass the unlabeled data through this seed system to automatically generate transcriptions; (iii) en-large the initial dataset with the self-labeled data and retrain the speech recognition system. However, in order not to pol-lute the augmented dataset with incorrect transcriptions, an important intermediary step is to select those parts of the self-labeled data that are accurate. Several approaches have been proposed in the community, but most of the works address only a single method. In contrast, in this paper we inspect three distinct classes of data-filtering for self-training, leveraging: (i) confidence scores, (ii) multiple ASR hypotheses and (iii) approximate transcriptions. We evaluate these approaches from two perspectives: quantity vs. quality of the selected data and improvement of the seed ASR by including this data. The proposed methodology achieves state-of-the-art results on Romanian speech, obtaining 25% relative improvement over prior work. Among the three methods, approximate transcriptions bring the highest performance gain, even if they yield the least quantity of data.","PeriodicalId":243211,"journal":{"name":"2021 IEEE Spoken Language Technology Workshop (SLT)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122806318","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: 3

Dual-Path RNN for Long Recording Speech Separation 用于长录音语音分离的双路径RNN

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2021-01-19 DOI: 10.1109/SLT48900.2021.9383514

Chenda Li, Yi Luo, Cong Han, Jinyu Li, Takuya Yoshioka, Tianyan Zhou, Marc Delcroix, K. Kinoshita, Christoph Böddeker, Y. Qian, Shinji Watanabe, Zhuo Chen

引用次数: 16

SLT 2021 Copyright Page SLT 2021版权页面

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2021-01-19 DOI: 10.1109/slt48900.2021.9383610

引用次数: 0

End-to-End Speaker Diarization Conditioned on Speech Activity and Overlap Detection 基于语音活动和重叠检测的端到端说话人分割

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2021-01-19 DOI: 10.1109/SLT48900.2021.9383555

Yuki Takashima, Yusuke Fujita, Shinji Watanabe, Shota Horiguchi, Leibny Paola García-Perera, Kenji Nagamatsu

引用次数: 19

Domain Generalization with Triplet Network for Cross-Corpus Speech Emotion Recognition 面向跨语料库语音情感识别的三元网络领域泛化

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2021-01-19 DOI: 10.1109/SLT48900.2021.9383534

Shi-wook Lee

引用次数: 8

An Investigation into the Multi-channel Time Domain Speaker Extraction Network 多通道时域说话人提取网络的研究

2021 IEEE Spoken Language Technology Workshop (SLT) Pub Date : 2021-01-19 DOI: 10.1109/SLT48900.2021.9383582

Catalin Zorila, Mohan Li, R. Doddipatla

{"title":"An Investigation into the Multi-channel Time Domain Speaker Extraction Network","authors":"Catalin Zorila, Mohan Li, R. Doddipatla","doi":"10.1109/SLT48900.2021.9383582","DOIUrl":"https://doi.org/10.1109/SLT48900.2021.9383582","url":null,"abstract":"This paper presents an investigation into the effectiveness of spatial features for improving time-domain speaker extraction systems. A two-dimensional Convolutional Neural Network (CNN) based encoder is proposed to capture the spatial information within the multichannel input, which are then combined with the spectral features of a single channel extraction network. Two variants of target speaker extraction methods were tested, one which employs a pre-trained i-vector system to compute a speaker embedding (System A), and one which employs a jointly trained neural network to extract the embeddings directly from time domain enrolment signals (System B). The evaluation was performed on the spatialized WSJ0-2mix dataset using the Signal-to-Distortion Ratio (SDR) metric, and ASR accuracy. In the anechoic condition, more than 10 dB and 7 dB absolute SDR gains were achieved when the 2-D CNN spatial encoder features were included with Systems A and B, respectively. The performance gains in reverberation were lower, however, we have demonstrated that retraining the systems by applying dereverberation preprocessing can significantly boost both the target speaker extraction and ASR performances.","PeriodicalId":243211,"journal":{"name":"2021 IEEE Spoken Language Technology Workshop (SLT)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130847315","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