HKUST/MTS: A Very Large Scale Mandarin Telephone Speech Corpus

By Yi Y. Liu, Pascale Fung, Yongsheng Yang, C. Cieri, Shudong Huang, David Graff. 2006

122 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Speech & Audio

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Speech & Audio

Citations in the corpus

Improving Transformer-based Speech Recognition Using Unsupervised Pre-training (2019, methodology): “The corpora we collected include: HKUST Mandarin Telephone Speech Corpus (HKUST/MTS) [18], AISHELL-1 [21], aidatatang 200zh, MAGICDATA Mandarin Chinese Read Speech Corpus, Free ST Chinese Mandarin Corpus (STCMDS) and Primewords Chinese Corpus Set 1.”
A Survey on Neural Speech Synthesis (2021, neutral): “1 English CSMSC [15] 12 1 48 Mandarin HKUST [205] 200 2100 8 Mandarin AISHELL-1 [28] 170 400 16 Mandarin AISHELL-2 [69] 1000 1991 44.”
A Comparison of Modeling Units in Sequence-to-Sequence Speech Recognition with the Transformer on Mandarin Chinese (2018, background): “The HKUST corpus (LDC2005S15, LDC2005T32), a corpus of Mandarin Chinese conversational telephone speech, is collected and transcribed by Hong Kong University of Science and Technology (HKUST) [16], which contains 150-hour speech, and 873 calls in the training set and 24 calls in the test set.”
ESPnet: End-to-End Speech Processing Toolkit (2018, methodology): “Although most of ASR recipes supported in ESPnet are standard English tasks, current ESPnet recipes deal with other languages including Japanese (CSJ), Mandarin Chinese (HKUST CTS), and other European languages through VoxForge.”
A Comparative Study on Transformer vs RNN in Speech Applications (2019, background): “For example, we applied speed perturbation [27] at ratio 0.9, 1.0 and 1.1 to CSJ, CHiME4, Fisher-CALLHOME Spanish, HKUST, and TED-LIUM2/3, and we also applied SpecAugment [26] to Aurora4, LibriSpeech, TED-LIUM2/3 and WSJ.2”
Extending Recurrent Neural Aligner for Streaming End-to-End Speech Recognition in Mandarin (2018, methodology): “At ﬁrst, we experiment with the Mandarin Chinese ASR benchmark (HKUST) [11] to investigate our four extensions in order.”
Didispeech: A Large Scale Mandarin Speech Corpus (2020, background): “[21] Y. Liu, P. Fung, Y. Yang et al., “HKUST/MTS: A very large scale mandarin telephone speech corpus,” in Proc.”
Multilingual Speech Recognition with Corpus Relatedness Sampling (2019, background): “English TED (ted) [22] broadcast 100,000 Mandarin Hkust (hk) [23] telephone 100,000 English Switchboard (swbd)[24] telephone 100,000 Mandarin SLR18 (s18) [25] read 13,388 English Librispeech (libri) [26] read 100,000 Mandarin LDC98S73 (hub) broadcast 35,999 English Fisher (fisher) [27] telephone 100”
Comparable Study Of Modeling Units For End-To-End Mandarin Speech Recognition (2018, methodology): “We perform the experiments on HKUST dataset [24], Callcenter dataset and Reading dataset, which are different not only on the the data size but also the dialogue scene.”
Constructing long short-term memory based deep recurrent neural networks for large vocabulary speech recognition (2014, methodology): “We evaluate these LSTM networks on a large vocabulary speech recognition task - the HKUST Mandarin Chinese conversation l telephone speech recognition [27].”
Multidimensional Residual Learning Based on Recurrent Neural Networks for Acoustic Modeling (2016, methodology): “On the HKUST task, unidirectional architecture “LSTMP3L-800C” and bidirectional architecture “BiLSTMP-3L-800C” are conducted for comparison with the following experiments.”
Advances in Joint CTC-Attention Based End-to-End Speech Recognition with a Deep CNN Encoder and RNN-LM (2017, background): “Tables 1 and 2 show character error rates (CERs) of evaluated methods in CSJ and MTS tasks, respectively.”
Modeling speaker variability using long short-term memory networks for speech recognition (2015, methodology): “We evaluate these networks on the HKUST Mandarin Chinese conversational telephone corpus (LDC2005S15, LDC2005T32) [34], which is collected and transcribed by Hong Kong University of Science and Technology (HKUST).”
Speech Recognition via Large-Scale Weak Supervision (unknown): “In Chinese, for example, the HKUST Mandarin Telephone Speech corpus has 1206 transcribed HKUST ten-minute telephone conversations between speakers of Mandarin across China including conversations between friends and between strangers (Liu et al., 2006).”
Large-Scale End-to-End Multilingual Speech Recognition and Language Identification with Multi-Task Learning (2020, background): “It is from 11 databases including: AISHELL [20], Au-rora4, Babel, CHiME4, Common Voice [21], Corpus of Spontaneous Japanese (CSJ) [22], Fisher Switchboard, Fisher Call-home Spanish, HKUST [23], WSJ and Voxforge.”
Hybrid CTC/Attention Architecture for End-to-End Speech Recognition (2017, methodology): “HKUST Mandarin Chinese conversational telephone speech recognition [37] has 5 hours of recording for evaluation, and we extracted an additional 5 hours from the training data as a development set, and used the rest (167 hours) as a training set, as summarized in Talbe I.”