JNAS: Japanese speech corpus for large vocabulary continuous speech recognition research.

Itou, Katunobu; Yamamoto, Michiya; Takeda, Kazuya; Takezawa, Toshiyuki; Matsuoka, Tatsuo; Kobayashi, Tetsuo; Shikano, Kiyohiro; Itahashi, Shuichi

doi:10.1250/ast.20.199

Cited by 214 publications

(85 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The frame length was 21.3 ms, and the frame shift was 8 ms with a 256-point Hamming window. Then, 116 Japanese speaker-independent syllable-based HMMs (strictly speaking, mora-unit HMMs [23]) were trained using 27,992 utterances read by 175 male speakers (JNAS corpus [24]). Each continuous-density HMM had five states, four with probability density functions (pdfs) of output probability.…”

Section: Methodsmentioning

confidence: 99%

Distant-Talking Speech Recognition Based on Spectral Subtraction by Multi-Channel LMS Algorithm

Wang

Kitaoka

Nakagawa

2011

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARYWe propose a blind dereverberation method based on spectral subtraction using a multi-channel least mean squares (MCLMS) algorithm for distant-talking speech recognition. In a distant-talking environment, the channel impulse response is longer than the short-term spectral analysis window. By treating the late reverberation as additive noise, a noise reduction technique based on spectral subtraction was proposed to estimate the power spectrum of the clean speech using power spectra of the distorted speech and the unknown impulse responses. To estimate the power spectra of the impulse responses, a variable step-size unconstrained MCLMS (VSS-UMCLMS) algorithm for identifying the impulse responses in a time domain is extended to a frequency domain. To reduce the effect of the estimation error of the channel impulse response, we normalize the early reverberation by cepstral mean normalization (CMN) instead of spectral subtraction using the estimated impulse response. Furthermore, our proposed method is combined with conventional delay-andsum beamforming. We conducted recognition experiments on a distorted speech signal simulated by convolving multi-channel impulse responses with clean speech. The proposed method achieved a relative error reduction rate of 22.4% in relation to conventional CMN. By combining the proposed method with beamforming, a relative error reduction rate of 24.5% in relation to the conventional CMN with beamforming was achieved using only an isolated word (with duration of about 0.6 s) to estimate the spectrum of the impulse response.

show abstract

Section: Methodsmentioning

confidence: 99%

Distant-Talking Speech Recognition Based on Spectral Subtraction by Multi-Channel LMS Algorithm

Wang

Kitaoka

Nakagawa

2011

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

show abstract

“…The duration for training and testing sets are about 42 hours and 5.3 hours. The Japanese corpus used for training is JNAS corpus [44], which is also a commonly used database for Japanese largevocabulary continuous speech recognition research. We randomly select 42 hours speech data (80 males and 73 females) from it.…”

Section: Databasementioning

confidence: 99%

Articulatory Modeling for Pronunciation Error Detection without Non-Native Training Data Based on DNN Transfer Learning

Duan

Kawahara

Dantsuji

et al. 2017

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARYAiming at detecting pronunciation errors produced by second language learners and providing corrective feedbacks related with articulation, we address effective articulatory models based on deep neural network (DNN). Articulatory attributes are defined for manner and place of articulation. In order to efficiently train these models of non-native speech without such data, which is difficult to collect in a large scale, several transfer learning based modeling methods are explored. We first investigate three closely-related secondary tasks which aim at effective learning of DNN articulatory models. We also propose to exploit large speech corpora of native and target language to model inter-language phenomena. This kind of transfer learning can provide a better feature representation of nonnative speech. Related task transfer and language transfer learning are further combined on the network level. Compared with the conventional DNN which is used as the baseline, all proposed methods improved the performance. In the native attribute recognition task, the network-level combination method reduced the recognition error rate by more than 10% relative for all articulatory attributes. The method was also applied to pronunciation error detection in Mandarin Chinese pronunciation learning by Japanese native speakers, and achieved the relative improvement up to 17.0% for detection accuracy and up to 19.9% for F-score, which is also better than the lattice-based combination.

show abstract

“…In this paper, we calculated the average speech spectrum and weight function by using JNAS database [9] with manual phoneme labeling. Figures 2 and 3 show an average speech spectrum and weight function, respectively.…”

Section: Design Of Weight Functionmentioning

confidence: 99%

Multiple-nulls-steering beamformer based on both talker and noise direction-of-arrival estimation

Nakayama

Nishiura

Yamashita

et al. 2013

Acoust. Sci. & Tech.

View full text Add to dashboard Cite

Beamforming with a microphone-array is an ideal candidate for distant-talking speech recognition. An adaptive beamformer can achieve beamforming with a small microphone-array, but it had difficulty extracting distant-moving speech and reducing moving noises, because it must rapidly train long multiple-channel adaptive filters by using observed noises with a microphone-array. However, if positions of both talkers and noises can be estimated, adaptive filters may not need to be trained in real noisy environments. Therefore, we propose a multiple-nulls-steering beamformer based on both talker and noise localization that does not require adaptive training with observed noises. Finally, we confirmed the validity and effectiveness of the proposed method through computer simulations and evaluation experiments in real noisy environments.

show abstract

JNAS: Japanese speech corpus for large vocabulary continuous speech recognition research.

Cited by 214 publications

References 3 publications

Distant-Talking Speech Recognition Based on Spectral Subtraction by Multi-Channel LMS Algorithm

Distant-Talking Speech Recognition Based on Spectral Subtraction by Multi-Channel LMS Algorithm

Articulatory Modeling for Pronunciation Error Detection without Non-Native Training Data Based on DNN Transfer Learning

Multiple-nulls-steering beamformer based on both talker and noise direction-of-arrival estimation

Contact Info

Product

Resources

About