Improving Semi-supervised Deep Neural Network. for Keyword Search in Low Resource Languages

被引：0

作者：

Hsiao, Roger ^{[1
]}

Ng, Tim ^{[1
]}

Le Zhang ^{[1
]}

Ranjan, Shivesh ^{[1
]}

机构：

[1] Raytheon BBN Technol, 10 Moulton St, Cambridge, MA 02138 USA

来源：

15TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2014), VOLS 1-4 | 2014年

关键词：

semi-supervised training; deep neural network; keyword search;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In this work, we investigate how to improve semi-supervised DNN for low resource languages where the initial systems may have high error rate. We propose using semi-supervised MLP features for DNN training, and we also explore using confidence to improve semi-supervised cross entropy and sequence training. The work conducted in this paper was evaluated under the IARPA Babel program for the keyword spotting tasks. We focus on the limited condition where there are around 10 hours of supervised data for training.

引用

页码：1088 / 1091

页数：4

共 50 条

[41] The Kaldi OpenKWS System: Improving Low Resource Keyword Search
Trmal, Jan
Wiesner, Matthew
Peddinti, Vijayaditya
Zhang, Xiaohui
Ghahremani, Pegah
Wang, Yiming
Manohar, Vimal
Xu, Hainan
Povey, Daniel
Khudanpur, Sanjeev
18TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2017), VOLS 1-6: SITUATED INTERACTION, 2017, : 3597 - 3601
[42] Deep Neural Backdoor in Semi-Supervised Learning: Threats and Countermeasures
Yan, Zhicong
Wu, Jun
Li, Gaolei
Li, Shenghong
Guizani, Mohsen
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, 2021, 16 : 4827 - 4842
[43] SEMI-SUPERVISED TRAINING IN LOW-RESOURCE ASR AND KWS
Metze, Florian
Gandhe, Ankur
Miao, Yajie
Sheikh, Zaid
Wang, Yun
Xu, Di
Zhang, Hao
Kim, Jungsuk
Lane, Ian
Lee, Won Kyum
Stueker, Sebastian
Mueller, Markus
2015 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING (ICASSP), 2015, : 4699 - 4703
[44] Deep Residual Spiking Neural Network for Keyword Spotting in Low-Resource Settings
Yang, Qu
Liu, Qi
Li, Haizhou
INTERSPEECH 2022, 2022, : 3023 - 3027
[45] On the Importance of Stereo for Accurate Depth Estimation: An Efficient Semi-Supervised Deep Neural Network Approach
Smolyanskiy, Nikolai
Kamenev, Alexey
Birchfield, Stan
PROCEEDINGS 2018 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS (CVPRW), 2018, : 1120 - 1128
[46] Semi-Supervised Deep Neural Network for Joint Intensity Estimation of Multiple Facial Action Units
Zhang, Yong
Fan, Yanbo
Dong, Weiming
Hu, Bao-Gang
Ji, Qiang
IEEE ACCESS, 2019, 7 : 150743 - 150756
[47] A combined deep neural network and semi-supervised clustering method for sports injury risk prediction
Zhao, Jiayi
Li, Guangxue
ALEXANDRIA ENGINEERING JOURNAL, 2023, 80 (191-201) : 191 - 201
[48] Hyperspectral Change Detection Using Semi-Supervised Graph Neural Network and Convex Deep Learning
Lin, Tzu-Hsuan
Lin, Chia-Hsiang
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2023, 61
[49] Semi-supervised text classification with deep convolutional neural network using feature fusion approach
Shayegh, Parvaneh
Li, Yuefeng
Zhang, Jinglan
Zhang, Qing
2019 IEEE/WIC/ACM INTERNATIONAL CONFERENCE ON WEB INTELLIGENCE (WI 2019), 2019, : 363 - 366
[50] Semi-supervised seismic attenuation compensation method based on GRU dominated deep neural network
Qi QianYu
Gao JingHuai
Chen HongLing
Gao ZhaoQi
Huang Yan
Chen Juan
Cao YongLiang
Wang JianHua
CHINESE JOURNAL OF GEOPHYSICS-CHINESE EDITION, 2023, 66 (07): : 2997 - 3010

← 1 2 3 4 5 →