研究者詳細 - 水町　光徳

論文 - 水町　光徳

分割表示195 件中 41 - 60 件目／全件表示 >>

Study of Ultrasonic Data Collection System for Anomaly Detection for Mechanical Equipment 査読有り国際誌

Shigyo R., Nobayashi D., Tsukamoto K., Mizumachi M., Ikenaga T.

Proceedings - 2022 International Conference on Computational Science and Computational Intelligence, CSCI 2022 1237 - 1238 2022年01月

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）

loT-based acoustic data collection systems have been proposed for detecting anomalies in equipment. However, such systems do not work well in outdoor environments due to interference caused by various sound sources (e.g., wildlife sounds and traffic noise). Outdoor noise has unique frequency characteristics, with energy localized in the low-frequency domain. In contrast, mechanical equipment such as motors mainly generates harmonics, including ultrasonic components, during operation. This paper proposes a noise-robust ultrasonic data collection system for outdoor use. To reduce the size of collected data, which increases in proportion to the sampling frequency, the proposed system uses bandpass filtering and data compression.

DOI： 10.1109/CSCI58124.2022.00222

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85172008371&origin=inward

公共空間の音声明瞭度を改善するEGG解析に基づく話者指標の検討

程島奈緒, 榊原健一, 水町光徳

日本音響学会春季研究発表会講演論文集 - 1123 - 1124 2022年01月

雑音下と残響下での発話におけるEGG解析による声帯振動様式の比較

程島奈緒, 榊原健一, 水町光徳

日本音響学会秋季研究発表会講演論文集 - 1615 - 1616 2022年01月

EGG解析から見た緊迫音声の声帯振動様式

程島奈緒, 榊原健一, 水町光徳

日本音響学会聴覚研究会資料 52(8) 625 - 630 2022年01月

Hardware Development of Color Image Contrast Enhancement by High-level Synthesis 査読有り

Yamawaki A., Mizumachi M., Serikawa S.

TENSYMP 2021 - 2021 IEEE Region 10 Symposium 2021年08月

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）

To realize image processing as hardware module is an effective way to achieve high performance and low power processing. However, hardware development is a heavy burden for software engineer handling image processing usually. High-level synthesis, HLS, converting software program to hardware automatically is a desirable technology for software engineer with less skill of hardware design. On the other hand, the HLS has a significant problem such as it tends to generate a large and low performance hardware when the software program input has never consideration of hardware organization. This paper, focusing on color image contrast enhancement based on histogram equalization, investigates software description methods considering hardware to make HLS generate a good hardware module with high-performance and low power.

DOI： 10.1109/TENSYMP52854.2021.9550810

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85117467339&origin=inward

Non-linear beamformer with long short-term memory network 査読有り国際誌

M. Mizumachi, R. Oka

Proc. Inter-noise 2021 2021年08月

Power-saving Reproduction Algorithm for Speech Signals 査読有り

M. Nakagawara, M. Mizumachi

Journal of the Institute of Industrial Applications Engineers 9 ( 3 ) 97 2021年07月

Performance evaluation of the nonlinear beamformer with the small-scall mems microphone array 査読有り国際誌

M. Mizumachi, R. Oka

Proc. 27th International Congress on Sound and Vibration (ICSV27) 2021年07月

Stereo Matching Based on Features of Image Patch 査読有り

S. Zhou, H. Li, M. Zhu, Z. Li, M. Mizumachi, L. Zhang

Journal of the Institute of Industrial Applications Engineers 9 ( 3 ) 91 2021年07月

Self-supervised monocular depth estimation with occlusion mask and edge awareness 査読有り国際誌

S. Zhou, M. Zhu, Z. Li, H. Li, M. Mizumachi, L. Zhang

Artificial Life and Robotics 26 ( 3 ) 354 - 359 2021年05月

記述言語：英語掲載種別：研究論文（学術雑誌）

Depth estimation is one of the basic and important tasks in 3D vision. Recently, many works have been done in self-supervised depth estimation based on geometric consistency between frames. However, these research works still have difficulties in ill-posed areas, such as occlusion areas and texture-less areas. This work proposed a novel self-supervised monocular depth estimation method based on occlusion mask and edge awareness to overcome these difficulties. The occlusion mask divides the image into two classes, making the training of the network more reasonable. The edge awareness loss function is designed based on the edge obtained by the traditional method, so that the method has strong robustness to various lighting conditions. Furthermore, we evaluated the proposed method on the KITTI datasets. The occlusion mask and edge awareness are both beneficial to find corresponding points in ill-posed areas.

DOI： 10.1007/s10015-021-00685-z

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85106505382&origin=inward

Perceptual optimization of stereo width control methods in headphones 査読有り国際誌

Y. Ueno, M. Mizumachi, T. Horiuchi

Proc. 150th AES Convention ( Audio Engineering Society ) Paper Number: 10480 2021年05月

Proposal of Speed-Up Method for Pulse Type of Digital Switch and Expansion of Application Scope to Them 招待有り査読有り

R. Iwaizono, M. Mizumachi, S. Serikawa

Proc. ICIAE 2021 VS-6 2021年03月

Proposal of a Digital Switch for Easy Changes of Number and Types 招待有り査読有り

R. Iwaizono, M. Mizumachi, S. Serikawa

Proc. ICIAE 2021 VS-3 2021年03月

[Keynote Lecture] Acoustic Signal Extraction Relying on Spatial Cues 招待有り

M. Mizumachi

Proc. ICIAE 2021 KS-1, Keynote Lecture 2021年03月

Development of a Wireless Communication System for Reliable Acoustic Data Collection Toward Anomaly Detection in Mechanical Equipment 査読有り国際誌

Takano C., Fujino S., Nobayashi D., Tsukamoto K., Mizumachi M., Ikenaga T.

Proceedings - 2021 International Conference on Computational Science and Computational Intelligence, CSCI 2021 1544 - 1547 2021年01月

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）

With the recent proliferation of Internet of Things (IoT) devices that can send and receive data via wireless communication, we are able to monitor and operate these devices remotely. An example of an IoT system using wireless communication is a system for anomaly detection in mechanical equipment using acoustic data. In order to detect anomalies using acoustic data, continuous recording is essential, thereby increasing the data size. Although Wi-Fi networks provide high-capacity data transfer, performance degradation cannot be avoided due to reasons such as packet losses caused by collisions with data from other devices using the same frequency and the increase in distance between two communicating devices. In the present study, we developed a wireless communication system for reliable acoustic data collection for anomaly detection in mechanical equipment. First, as preliminary experiments, we investigated the communication characteristics for the transmission of large-size data by Wi-Fi in indoor and outdoor environments. The results indicated the communication performance was insufficient for transferring all recorded data handled by this system. Therefore, we developed a simple heuristic transmission timing control method and a method that can reduce the amount of transmission data in order to realize a stable acoustic data collection system. Finally, through demonstration experiments using mechanical equipment in the field, we verified the feasibility of the acoustic data collection system.

DOI： 10.1109/CSCI54926.2021.00300

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85133931264&origin=inward

Sharpening Directivity of Beamforming based on Neural Networks using Virtual Observations 査読有り

Kanekiyo Y., Mizumachi M., Miyazaki R.

2021 IEEE 10th Global Conference on Consumer Electronics, GCCE 2021 512 - 513 2021年01月

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）

Beamforming based on neural networks (NN) has been proposed as nonlinear beamforming. This method improves the performance with more microphones, but it may not be possible to easily increase the number of microphones due to hardware restrictions. Therefore, we focus on virtual observation, which is an estimation of the signal observed at a location where the microphone is not actually placed, using only actual observations. In this paper, we propose to apply virtual observation to NN-based beamforming. From the evaluation experiment, we confirmed that the proposed method can obtain sharper directivity than the case of using only actual microphones.

DOI： 10.1109/GCCE53005.2021.9621789

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85123489127&origin=inward

Performance evaluation of the nonlinear beam-former with small-scale mems microphone array 査読有り国際誌

Mizumachi M., Oka R.

"Advances in Acoustics, Noise and Vibration - 2021" Proceedings of the 27th International Congress on Sound and Vibration, ICSV 2021 2021年01月

担当区分：筆頭著者,　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）

Beamforming has been one of the most important techniques in noise reduction under adverse acoustic environments. Traditional beamformers were designed in analytical and adaptive schemes. On the other hand, recent neural network-based beamformers achieve great success in signal enhancement. The authors have also proposed a broadband neural network-based beamformer, which can deal with wideband speech signals. The proposed beamformer employs the long short-term memory to model the temporal relationship among multichannel observations. The performance of the proposed beamformer has been confirmed with an 8-ch small-scale linear microphone array, of which the spacing between neighbouring microphones is 10 mm. It is confirmed that the proposed nonlinear beamformer could sharpen the mainlobe and decrease the sidelobes, although an analytical linear beamformer, that is, delay-and-sum beamformer, could not achieve signal enhancement with the small-scale microphone array.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85117473124&origin=inward

Perceptual optimization of stereo width control methods in headphones 査読有り国際誌

Ueno Y., Mizumachi M., Horiuchi T.

150th Audio Engineering Society Convention, AES 2021 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）

Legacy 2-ch stereo music sources cause unnatural spatial impressions through earphones and headphones reproduction due to the lack of the crosstalk produced in loudspeakers reproduction. Perceptual widths of the stereo stage in amplitude-based and phase-based stereo width control methods for headphones reproduction were matched with that of the binaural rendering which simulates loudspeakers reproduction. Paired comparison between the stereo width control methods and binaural rendering was carried out to achieve the perceptual optimization for classic, jazz, and pop music sources. Concerning the amplitude-based stereo width control method, the suitable parameter value for each sound source could be determined from the results of the listening test. It is also found that the phase-based stereo width control method should be optimized in different parameter spaces.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85113887802&origin=inward

Non-linear beamformer with long short-term memory network 査読有り国際誌

Mizumachi M., Oka R.

Proceedings of INTER-NOISE 2021 - 2021 International Congress and Exposition of Noise Control Engineering 2021年01月

担当区分：筆頭著者,　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）

Acoustic beamforming with a microphone array enables spatial filtering in a wide frequency range. It is a challenging issue to sharpen the mainlobe in the lower frequency region with a small-scale microphone array, of which the number and spacing of microphones are small. A neural network-based non-linear beamformer achieves a breakthrough in sharpening the main-lobe. The non-linear beamforming works well for the narrowband signals but is weak in wideband beamforming. The nonlinear beamforming with the long short-term memory is proposed to deal with wideband speech signals. The long short-term memory network is trained in the recurrent neural network architecture with the sequence of audio data such as speech signals. The performance of the proposed beamformer is confirmed using a small-scale 8-ch MEMS microphone array, where eight microphones are linearly arranged with the neighboring spacing of 10 mm, under a real environment. The beam-pattern of the proposed non-linear beamformer succeeds in sharpening the main-lobe although the linear delay-and-sum beamformer could not achieve frequency selectivity. The feasibility of the proposed beamformer is also confirmed in speech enhancement.

DOI： 10.3397/IN-2021-2673

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85117405196&origin=inward

Interactive and real-time acoustic measurement tools for speech data acquisition and presentation: Application of an extended member of time stretched pulses 査読有り

Kawahara H., Yatabe K., Sakakibara K.I., Mizumachi M., Morise M., Banno H., Irino T.

Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH 3 2197 - 2198 2021年01月

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）

Objective measurements of speech data acquisition and presentation processes are crucial for assuring reproducibility and reusability of experimental results and acquired materials. We introduce setting and measurement examples of those conditions using an interactive and real-time acoustic measurement tool based on an extended time-stretched pulse. We also introduce supporting tools.

Scopus

その他リンク： https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85119206055&origin=inward

　前のページ - 次のページ　

<水町 光徳>

論文 - 水町 光徳

<水町　光徳>

論文 - 水町　光徳