2,508 Hits in 6.5 sec

Unsupervised Domain Adaptation by Adversarial Learning for Robust Speech Recognition [article]

Pavel Denisov, Ngoc Thang Vu, Marc Ferras Font
2018 arXiv   pre-print
In this paper, we investigate the use of adversarial learning for unsupervised adaptation to unseen recording conditions, more specifically, single microphone far-field speech.  ...  We adapt neural networks based acoustic models trained with close-talk clean speech to the new recording conditions using untranscribed adaptation data.  ...  Crosslingual Adaptation Conclusions The present study was designed to gain a better understanding of ability of unsupervised domain adaptation by adversarial Learning to improve robustness of ASR.  ... 
arXiv:1807.11284v1 fatcat:zurxttzyejf67gdrlnzh7wu4wy

Domain Adversarial Training for Accented Speech Recognition [article]

Sining Sun, Ching-Feng Yeh, Mei-Yuh Hwang, Mari Ostendorf, Lei Xie
2018 arXiv   pre-print
Furthermore, we find that DAT is superior to multi-task learning for accented speech recognition.  ...  In this paper, we propose a domain adversarial training (DAT) algorithm to alleviate the accented speech recognition problem.  ...  Among these, domain adaptation is also of great interest for robust speech recognition, especially for DL-based methods.  ... 
arXiv:1806.02786v1 fatcat:cmdie6p2jngblhyejfolefl2qy

Unsupervised adaptation with domain separation networks for robust speech recognition

Zhong Meng, Zhuo Chen, Vadim Mazalov, Jinyu Li, Yifan Gong
2017 2017 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)  
training method, for automatic speech recognition on CHiME-3 dataset.  ...  Unsupervised domain adaptation of speech signal aims at adapting a well-trained source-domain acoustic model to the unlabeled data from target domain.  ...  In this work, we propose to apply DSN for unsupervised domain adaptation on a DNN-hidden Markov model (HMM) acoustic model, aiming to increase the noise robustness in speech recognition.  ... 
doi:10.1109/asru.2017.8268938 dblp:conf/asru/MengCMLG17 fatcat:i2h33gjzuzfynfyhrf5yv2vkse

Unsupervised Adversarial Domain Adaptation for Cross-Lingual Speech Emotion Recognition [article]

Siddique Latif, Junaid Qadir, Muhammad Bilal
2020 arXiv   pre-print
Cross-lingual speech emotion recognition (SER) is a crucial task for many real-world applications.  ...  Therefore, in this paper, we propose a Generative Adversarial Network (GAN)-based model for multilingual SER.  ...  CONCLUSIONS In this paper, we proposed an unsupervised adversarial domain adaption approach for developing deep learning models for cross-lingual speech emotion recognition tasks.  ... 
arXiv:1907.06083v4 fatcat:pbxexa5ydzdqlj63spueap5cga

Adversarial Teacher-Student Learning for Unsupervised Domain Adaptation

Zhong Meng, Jinyu Li, Yifan Gong, Biing-Hwang Juang
2018 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
In this work, we advance T/S learning by proposing adversarial T/S learning to explicitly achieve condition-robust unsupervised domain adaptation.  ...  The teacher-student (T/S) learning has been shown effective in unsupervised domain adaptation [1].  ...  To benefit from both methods, in this work, we advance T/S learning with adversarial T/S training for condition-robust unsupervised domain adaptation, where a student acoustic model and a domain classifier  ... 
doi:10.1109/icassp.2018.8461682 dblp:conf/icassp/MengLGJ18 fatcat:qj2osqass5anti5s5q7h54irti

Adversarial Multi-Task Learning of Deep Neural Networks for Robust Speech Recognition

Yusuke Shinohara
2016 Interspeech 2016  
A method of learning deep neural networks (DNNs) for noise robust speech recognition is proposed.  ...  In this paper, we propose adversarial multi-task learning of DNNs for explicitly enhancing the invariance of representations.  ...  [15, 16] for unsupervised domain adaptation, but its application to supervised learning tasks has never been examined before.  ... 
doi:10.21437/interspeech.2016-879 dblp:conf/interspeech/Shinohara16a fatcat:p7hrt75grjcx7cwvaaagywhqom

Unsupervised Adaptation with Adversarial Dropout Regularization for Robust Speech Recognition

Pengcheng Guo, Sining Sun, Lei Xie
2019 Interspeech 2019  
Recent adversarial methods proposed for unsupervised domain adaptation of acoustic models try to fool a specific domain discriminator and learn both senone-discriminative and domaininvariant hidden feature  ...  Thus, ambiguous target domain features can be generated near the decision boundaries, decreasing speech recognition performance.  ...  Acknowledgements This research work is supported by the National Natural Science Foundation of China (No.61571363).  ... 
doi:10.21437/interspeech.2019-2544 dblp:conf/interspeech/GuoS019 fatcat:twkl6br2cbgt7dkhq6vgsl6rua

Senone-aware Adversarial Multi-task Training for Unsupervised Child to Adult Speech Adaptation [article]

Richeng Duan, Nancy F. Chen
2021 arXiv   pre-print
In this work, we propose a feature adaptation approach by exploiting adversarial multi-task training to minimize acoustic mismatch at the senone (tied triphone states) level between adult and child speech  ...  Acoustic modeling for child speech is challenging due to the high acoustic variability caused by physiological differences in the vocal tract.  ...  UNSUPERVISED FEATURE ADAPTATION VIA ADVERSARIAL MULTI-TASK LEARNING Feature adaptation with binary domain adversaries The idea of domain adversarial learning is minimizing the domain discriminator prediction  ... 
arXiv:2102.11488v1 fatcat:b4qbujxjv5gorgjoqn4cfzphma

A Multi-Discriminator CycleGAN for Unsupervised Non-Parallel Speech Domain Adaptation

Ehsan Hosseini-Asl, Yingbo Zhou, Caiming Xiong, Richard Socher
2018 Interspeech 2018  
Domain adaptation plays an important role for speech recognition models, in particular, for domains that have low resources.  ...  We propose a novel generative model based on cyclicconsistent generative adversarial network (CycleGAN) for unsupervised non-parallel speech domain adaptation.  ...  Therefore, an unsupervised domain adaptation is desirable for building a robust ASR system.  ... 
doi:10.21437/interspeech.2018-1535 dblp:conf/interspeech/Hosseini-AslZXS18 fatcat:d76yrtaunjcthldvvb4mt5ddfm

Deep Representation Learning in Speech Processing: Challenges, Recent Advances, and Future Trends [article]

Siddique Latif, Rajib Rana, Sara Khalifa, Raja Jurdak, Junaid Qadir, Björn W. Schuller
2021 arXiv   pre-print
distinct research areas including Automatic Speech Recognition (ASR), Speaker Recognition (SR), and Speaker Emotion Recognition (SER).  ...  Recent reviews in speech have been conducted for ASR, SR, and SER, however, none of these has focused on the representation learning from speech -- a gap that our survey aims to bridge.  ...  Also, some studies [182], [183] explored unsupervised representation learning-based domain adaptation for distant conversational speech recognition.  ... 
arXiv:2001.00378v2 fatcat:ysvljxylwnajrbowd3kfc7l6ve

Adversarial Learning of Raw Speech Features for Domain Invariant Speech Recognition [article]

Aditay Tripathi, Aanchan Mohan, Saket Anand, Maneesh Singh
2018 arXiv   pre-print
Promising empirical results indicate the strength of adversarial training for unsupervised domain adaptation in ASR, thereby emphasizing the ability of DANNs to learn domain invariant features from raw  ...  This paper explores the application of adversarial training to learn features from raw speech that are invariant to acoustic variability.  ...  Sun [16] arXiv:1805.08615v1 [eess.AS] 21 May 2018 used adversarial training for unsupervised domain adaptation for robust speech recognition using filter bank features and used the WSJ and Librispeech  ... 
arXiv:1805.08615v1 fatcat:ehbdqulelfdgbk4yqtlfitvbau

Cross-lingual Text-independent Speaker Verification Using Unsupervised Adversarial Discriminative Domain Adaptation

Wei Xia, Jing Huang, John H.L. Hansen
2019 ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
In this study, we introduce an unsupervised Adversarial Discriminative Domain Adaptation (ADDA) method to effectively learn an asymmetric mapping that adapts the target domain encoder to the source domain  ...  Further data analysis of ADDA adapted speaker embedding shows that the learned speaker embeddings can perform well on speaker classification for the target domain data, and are less dependent with respect  ...  SRE corpus is part of the Mixer 6 project, which was designed to support the development of robust speaker recognition technology by providing carefully collected speech across numerous microphones.  ... 
doi:10.1109/icassp.2019.8682259 dblp:conf/icassp/XiaHH19 fatcat:ulwrq5klbbad7dfbzdtg3b6sga

Unsupervised Feature Adaptation Using Adversarial Multi-Task Training for Automatic Evaluation of Children's Speech

Richeng Duan, Nancy F. Chen
2020 Interspeech 2020  
In this work, we tackle such challenges by proposing an unsupervised feature adaptation approach based on adversarial multi-task training in a neural framework.  ...  Experimental results demonstrate that our proposed approach consistently outperforms established baselines trained on adult speech across a variety of tasks ranging from speech recognition to pronunciation  ...  classification [16] , robust speech recognition [17] , speaker adaptation [18] , and spoken keyword spotting [19] .  ... 
doi:10.21437/interspeech.2020-1657 dblp:conf/interspeech/DuanC20 fatcat:fmyri7jmwbazfpsztsc3awihd4

Unsupervised Speech Domain Adaptation Based on Disentangled Representation Learning for Robust Speech Recognition [article]

Jong-Hyeon Park, Myungwoo Oh, Hyung-Min Park
2019 arXiv   pre-print
We improved word accuracies by 6.55~15.70\% for the CHiME4 challenge corpus by applying a noisy-to-clean environment adaptation for robust ASR.  ...  In this paper, we propose a domain adaptation method based on generative adversarial nets (GANs) with disentangled representation learning to achieve robustness in ASR systems.  ...  ) [11] was applied to learn a mapping for speech-tospeech adaptation from a speaker to a target speaker for gendermismatched recognition [12] .  ... 
arXiv:1904.06086v1 fatcat:qnak75aupfeqrd6q47n2ymmno4

Unsupervised Domain Adaptation in Speech Recognition using Phonetic Features [article]

Rupam Ojha, C Chandra Sekhar
2021 arXiv   pre-print
As a result, domain adaptation is important in speech recognition where we train the model for a particular source domain and test it on a different target domain.  ...  In this paper, we propose a technique to perform unsupervised gender-based domain adaptation in speech recognition using phonetic features.  ...  Approaches to explore the gender based domain adaptation for speech recognition include cycle-GAN [7] , multi-discriminator cycle-GAN [8] and augmented cycle adversarial learning [9] .  ... 
arXiv:2108.02850v1 fatcat:mp4prfyp4repvbjtaf2qwy7qmu
« Previous Showing results 1 — 15 out of 2,508 results