研究者詳細 - 中藤　良久

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84984829622&origin=inward

LPC残差駆動歌唱システムにおける音韻性除去による音質改善の検討査読有り

姫野剛至, 田中裕人, 水町光徳, 松井謙二, 中藤良久

産業応用工学会論文誌 ( 一般社団法人産業応用工学会 ) 4 ( 1 ) 19 - 25 2016年01月

担当区分：責任著者記述言語：日本語掲載種別：研究論文（学術雑誌）

We propose a singing system to transmit the pseudo-vocal sound source into the vocal tract. In this paper, we examine the pseudo-vocal sound source that is suitable for singing system. So we carried out a comparison of LPC residual wave and Bark-LPC residual wave that the phonological component included in pseudo-vocal sound source was removed by Bark spectrum. As a result, we found that although the singing voice using Bark-LPC residual is in the low natural tendencies, the Bark-LPC residual wave can reproduce a higher speaker information contributed by the speaker characteristics.

DOI： 10.12792/jjiiae.004.01.04

CiNii Research

その他リンク： http://ci.nii.ac.jp/naid/130005263509

Singing System by Electric-Larynx Using LPC Residual Wave 査読有り国際誌

Takashi Himeno, Yuto Tanaka, Mitsunori Mizumachi, Yoshihisa Nakatoh, Kenji Matsui

Journal of the Institute of Industrial Applications Engineers 3 ( 4 ) 183 - 186 2015年12月

Study of Hearing Aid System Adapted to Direction Inputted by User 査読有り

NOGUCHI TSUBASA, MIZUMACHI MITSUNORI, NAKATOH YOSHIHISA

Journal of the Institute of Industrial Applications Engineers ( The Institute of Industrial Applications Engineers ) 3 ( 2 ) 72 - 79 2015年04月

文字入力を用いた発話支援システムの開発査読有り

濱井裕也, 水町光徳, 中藤良久, 松井謙二

産業応用工学会論文誌 3 ( 1 ) 10 - 17 2015年01月

担当区分：責任著者記述言語：日本語掲載種別：研究論文（学術雑誌）

Utterance is very important for us to communicate with others. However, many people have speech handicaps and can't produce utterance smoothly. They mainly rely on writing and sign language for communication. But they are not easy skills for them to master. In order to help them, we developed a new utterance support system which can function through an information terminal. We made predictions which includes two additional functions of fixed phrase dictionary and the class 2-gram – one of the statistical approaches for processing natural language. Fixed phrase dictionary includes phrases often used in our daily life conversations and class 2-gram uses co-occurrence frequencies of parts of speech in Japanese. There are two purposes for us to use class 2-gram, one is to generate next candidates and the other is to change frequency. We concluded two evaluation experiments. The results show that our system needs less numbers of inputs in helping with written utterance production and using our prototype permits us to input more characters in one minute than using iPhone IME. Hereafter, we intend to create more effective dictionaries and develop new GUI to make our system more effective and faster.

DOI： 10.12792/jjiiae.003.01.02

その他リンク： http://ci.nii.ac.jp/naid/130005263536

Study of Improvement of Intelligibility for Elderly Speech Based on Formant Frequency Shift 査読有り国際誌

Yuto Tanaka, Hiroaki Igaue, Mitsunori Mizumachi, Yoshihisa Nakatoh

International Journal of Computer, Consumer and Control 3 ( 3 ) 57 - 65 2014年08月

音声認識応用システムのための音声コマンド設計ガイドライン査読有り

花沢健, 中藤良久, 中嶌信弥

電子情報通信学会誌 = The journal of the Institute of Electronics, Information and Communication Engineers ( 一般社団法人電子情報通信学会 ) 97 ( 7 ) 582 - 587 2014年07月

記述言語：日本語掲載種別：研究論文（学術雑誌）

近年,音声認識技術とそれを応用したシステムは急速に身近になってきている.音声認識応用システムでは,設計段階でアプリケーションの機能とそれに対応する音声コマンドを定義する必要がある.音声コマンドの設計は,開発者にとってはシステムの性能を引き出す重要事項であり,利用するユーザにとっては使い勝手を決定する主要因となる.JEITA音声入出力方式標準化専門委員会では,音声コマンドの親密度や正確性,発声や音声認識のしやすさなど,音声コマンドの設計における留意点や指針を整理し,ガイドラインとしてまとめている.本稿では,そのエッセンスについて解説する.

その他リンク： https://ci.nii.ac.jp/naid/110009830844

The guidelines for design of voice commands in application systems with speech recognition 査読有り国際誌

Hanazawa K., Nakatoh Y., Nakajima S.

Journal of the Institute of Electronics, Information and Communication Engineers 97 ( 7 ) 582 - 587 2014年01月

Study of Utterance Support System using Word Input based on Word Prediction 査読有り

HAMAI YUYA, MIZUMACHI MITSUNORI, NAKATOH YOSHIHISA

Journal of the Institute of Industrial Applications Engineers ( 産業応用工学会 ) 2 ( 1 ) 13 - 18 2014年01月

担当区分：責任著者記述言語：英語掲載種別：研究論文（学術雑誌）

Speech is very important for us to communicate with others. However, because of speech handicaps, there are many people who feel it difficult. Thus, we set our main purpose to study new utterance support system working on an information terminal. To make this system more effective, we made word prediction and added fixed phrase dictionary and class 2-gram as additional functions. Fixed phrase dictionary includes phrases often used in daily life conversations and class 2-gram using co-occurrence frequencies of 9 parts of speech in Japanese. And by using class 2-gram, we made two additional functions. One is generating next candidates and another is changing frequency. We did three evaluation experiments – necessary numbers of input, input speed and measurement experiments of voice synthesis. And results shows that our system needs less numbers of input but slower than other methods and it is faster to do voice synthesis when user input a sentence than other timing. Hereafter, we intend to create more effective dictionaries, improve class 2-gram and develop new GUI to make our system more effective and faster.

その他リンク： https://ci.nii.ac.jp/naid/120006870840

Speech enhancement by non-linear beamforming tolerant to misalignment of target source direction 査読有り

HAYASHI HIROAKI, MIZUMACHI MITSUNORI, NAKATOH YOSHIHISA

Journal of the Institute of Industrial Applications Engineers ( The Institute of Industrial Applications Engineers ) 1 ( 2 ) 97 - 104 2013年10月

Development of Hand Gesture Controlled Electrolarynx 査読有り

MATSUI KENJI, NAKATOH YOSHIHISA, KATO YUMIKO O.

Journal of the Institute of Industrial Applications Engineers ( The Institute of Industrial Applications Engineers ) 1 ( 2 ) 48 - 51 2013年10月

A low-bit-rate audio codec using mel-scaled linear predictive analysis 査読有り

Nakatoh Y., Matsumoto H.

Acoustical Science and Technology 28 ( 3 ) 147 - 152 2007年05月

担当区分：筆頭著者記述言語：英語掲載種別：研究論文（学術雑誌）

In this paper, we propose a low-bit-rate audio codec using a new analysis method named mel-scaled linear predictive analysis (mel-LP analysis). In mel-LP analysis, a spectral envelope is estimated on a mel- or bark-frequency scale, so as to improve the spectral resolution in the low-frequency band. This analysis is accomplished with about a twofold increase in computation over standard LPC analysis. Our codec using mel-LP analysis consists of five key parts: time frequency transformation, flattening of MDCT coefficients using the mel-LP spectral envelope, power normalization, perceptual weighting estimation, and multistage VQ. In subjective experiments, we investigated the performance of our codec using the mel-LP analysis method, through 7-level paired comparison tests. The result shows that the codec using the mel-LP analysis method results in a good performance at a low bit rate, particularly at 16 kbps. In the cases of pop songs, piano music and male speech, sound quality was improved. © 2007 The Acoustical Society of Japan.

DOI： 10.1250/ast.28.147

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=34247642253&origin=inward

Speech recognition interface system for digital TV control 査読有り

Nakatoh Y., Kuwano H., Kanamori T., Hoshimi M.

Acoustical Science and Technology 28 ( 3 ) 165 - 171 2007年05月

担当区分：筆頭著者記述言語：英語掲載種別：研究論文（学術雑誌）

In this paper, we describe a speech recognition interface system for digital TV (DTV) control. TV systems are currently undergoing digitalization and will become more multifunctional, leading to more complex TV operations. Thus, it is necessary for everyone to be able to use TVs easily, and a speech recognition interface is an important key technology. A speech recognition system, which is designed for home use, particularly for digital TV, must be simple and robust to environmental noises and speaker variations. To provide robustness to noise, we developed a noise reduction technique for house noise and an echo-canceling technique for TV sound. To achieve robustness to speaker variations, we developed new speaker adaptation techniques which are incorporated in the system. These of technologies results in a significant improvement in the recognition performance of the DTV. © 2007 The Acoustical Society of Japan.

DOI： 10.1250/ast.28.165

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=34247639555&origin=inward

Generation of broadband speech from narrowband speech based on linear mapping 査読有り

Nakatoh Y., Tsushima M., Norimatsu T.

Electronics and Communications in Japan, Part II: Electronics (English translation of Denshi Tsushin Gakkai Ronbunshi) 85 ( 8 ) 44 - 53 2002年08月

担当区分：筆頭著者記述言語：英語掲載種別：研究論文（学術雑誌）

In this paper, a method for generating broadband speech from band-limited speech using spectral linear mapping is proposed. This method is based on LPC analysis and synthesis. It first extracts sound source information (residual waveforms) and vocal tract information (spectral envelopes) from input speech, performs linear mapping of the vocal tract information and nonlinear processing of the sound source information for broadband conversion, and finally generates broadband speech from both by LPC synthesis. The spectral envelope is made broadband by linear mapping, the spectral space is divided into a number of subspaces, and a narrowband spectrum is converted into a broadband spectrum by a conversion (convert) matrix of each subspace. A conversion (convert) matrix is estimated using training speech such that the mean square error between the postconversion spectrum and the target broadband spectrum is minimized. Spectral distortions of the proposed method were compared experimentally with codebook mapping and neural network methods and it has been verified that the proposed method with linear mapping gives performance not inferior to that of the other two methods. In addition, the proposed method has been verified to have the effect of giving a band feeling in subject evaluation tests. © 2002 Wiley Periodicals, Inc. Electron. Comm. Jpn. Pt.

DOI： 10.1002/ecjb.10065

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=0036687129&origin=inward

スペストル線形写像による帯域制限音声の広帯域化査読有り

中藤良久, 津島峰生, 則松武志

電子情報通信学会論文誌. D-2, 情報・システム 2-パターン処理 ( 一般社団法人電子情報通信学会 ) 83 ( 11 ) 2246 - 2254 2000年11月

担当区分：筆頭著者記述言語：日本語掲載種別：研究論文（学術雑誌）

本論文では, スペクトル線形写像を用いて帯域制限された音声から広帯域音声を生成する方法について提案している.本方法はLPC分析合成を基本とし, まず入力音声から声道情報(スペクトル包絡)と音源情報(残差波形)を抽出し, 声道情報は線形写像により, また音源情報は非線形処理を施すことで広帯域化する.最後に, この両者からLPC合成により広帯域化音声を作成している.線形写像によるスペクトル包絡の広帯域化は, スペクトル空間をいくつかの部分空間に分割し, 各部分空間ごとの変換行列により狭帯域スペクトルから広帯域スペクトルへと変換する.変換行列は, 変換後のスペクトルと目標となる広帯域スペクトルとの間の平均2乗誤差を最小化するように学習用音声を用いて推定される.実験では, 本手法のほかにコードブック写像及びニューラルネットを用いる方法についてスペクトルひずみに関して比較した結果, 線形写像による方法は他の2手法に劣らない性能が得られることが確認できた.また主観評価実験によっても, 帯域感を付与する効果が確認できた.

その他リンク： https://ci.nii.ac.jp/naid/110003183835

Estimation of Road Damage Contents for Earthquake Evacuation Guidance System 査読有り国際誌

Yujiro Mihara, Rin Hirakawa, Hideaki Kawano, Kenichi Nakashi, Yoshihisa Nakatoh

Proc. of 3rd International Conference on Communication and Computational Technologies (ICCCT 2021) 2021年02月

Speech Enhancement System using SVM for Train Announcement 査読有り

Kinoshita Y., Hirakawa R., Kawano H., Nakashi K., Nakatoh Y.

Digest of Technical Papers - IEEE International Conference on Consumer Electronics 2021-January 2021年01月

担当区分：責任著者記述言語：英語掲載種別：研究論文（国際会議プロシーディングス）

On the train, announcement speech is one of the important information to know the operation status. However, clarity of announcement speech is not enough because announcement speech is masked by noise. Therefore, new system is proposed to improve clarity of announcement speech based on sound classification. By classifying several sounds, users can select the announcement speech only they want to hear. In this paper, classifier is developed that distinguishes announcement speech using SVM. In the experiments, this classifier obtain that the accuracy of developed classifier is 96%.

DOI： 10.1109/ICCE50685.2021.9427600

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85106052569&origin=inward

Eye Fatigue Prediction System Using Blink Detection Based on Eye Image 査読有り

Kuwahara A., Hirakawa R., Kawano H., Nakashi K., Nakatoh Y.

Digest of Technical Papers - IEEE International Conference on Consumer Electronics 2021-January 2021年01月

担当区分：責任著者記述言語：英語掲載種別：研究論文（国際会議プロシーディングス）

Today, information devices such as smartphones have become indispensable in our daily life. As a result, the increased use of information devices has led to the accumulation of eye strain and the development of various eye diseases. As the accumulation of eye-fatigue is manifested in the change in the number of blinks, blink detection technology is necessary. In the previous paper, we proposed a highly accurate blink detection method called EARM. However, there is a problem with this technique, which is that the noise generated by the movement of the face can reduce the accuracy of blink detection. In this study, we used face image normalization to improve the blink detection system by reducing the noise associated with face movement. We found that the proposed system outperformed the results of blink detection accuracy using EARM in all subjects.

DOI： 10.1109/ICCE50685.2021.9427681

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85106066793&origin=inward

Detecting System Alzheimer's Dementia by 1d CNN-LSTM in Japanese Speech 査読有り

Nishikawa K., Hirakawa R., Kawano H., Nakashi K., Nakatoh Y.

Digest of Technical Papers - IEEE International Conference on Consumer Electronics 2021-January 2021年01月

担当区分：責任著者記述言語：英語掲載種別：研究論文（国際会議プロシーディングス）

Dementia patients have increased year by year in many developed countries. We need to solve this problem. The questionnaire method is generally used for present dementia screening. However, this method requires time for 10 to 15 minutes with a doctor and a clinical psychologist, which puts a burden on hospitals and test subjects. For this reason, we propose a detecting method using deep learning by means of 1d CNN-LSTM to construct a dementia discriminant system using speech voice. Our method obtained the Accuracy of 0.908 and the F-measure of 0.897.

DOI： 10.1109/ICCE50685.2021.9427692

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85106001096&origin=inward

Blink detection using image processing to predict eye fatigue 査読有り

Kuwahara A., Hirakawa R., Kawano H., Nakashi K., Nakatoh Y.

Advances in Intelligent Systems and Computing 1253 AISC 362 - 368 2021年01月

担当区分：責任著者記述言語：英語掲載種別：研究論文（学術雑誌）

With the use of information terminals represented by smartphones, our eyes get tired. One index indicating eye fatigue is a change in the number of blinks. In this study, we set the ultimate goal that is to develop eye fatigue prevention system with less time and less physical burden for patient. By doing this, the information terminal performs a flickering of the eyes using a detection camera and a validity check. Previous EAR study with blink detection is confirmed to be effective, and this paper proposes a new formula EARM used EAR. As an evaluation method of blink detection, we used the total time average of the square of the residual value. The results showed that EARM was more accurate than the EAR. Further, it was suggested that the number of blinks during VDT work can be classified into several patterns.

DOI： 10.1007/978-3-030-55307-4_55