|
Affiliation Department |
電気・機械工学科 電気電子分野
|
|
External Link |
|
TANAKA Yuji
|
|
|
Research Areas
-
Life Science / Biomedical engineering / Physiological Signal Processing
-
Manufacturing Technology (Mechanical Engineering, Electrical and Electronic Engineering, Chemical Engineering) / Measurement engineering / Radar Signal Processing
From School
-
Kanazawa University Graduate School of Natural Science and Technology Division of Electrical Engineering and Computer Science Graduated
2020.04 - 2023.03
Notes：Doctoral Program
-
Kanazawa University Graduate School of Natural Science and Technology Division of Electrical Engineering and Computer Science Graduated
2015.04 - 2017.03
Notes：Master's Program
-
Kanazawa University College of Science and Engineering School of Electrical and Information Engineering Graduated
2011.04 - 2015.03
External Career
-
Kyoto University Graduate School of Engineering Department of Electrical Engineering Assistant Professor
2023.04 - 2024.03
Country：Japan
-
Information Technology R&D Center, Mitsubishi Electric Corporation
2017.04 - 2020.03
Professional Memberships
Papers
-
Deep Learning Model Size Performance Evaluation for Lightning Whistler Detection on Arase Satellite Dataset Reviewed International coauthorship International journal
I Made Agus Dwi Suarjaya, Desy Purnami Singgih Putri, Yuji Tanaka, Fajar Purnama, I Putu Agung Bayupati, Linawati, Yoshiya Kasahara, Shoya Matsuda, Yoshizumi Miyoshi, Iku Shinohara
Remote Sensing 16 ( 22 ) 2024.11
Language：English Publishing type：Research paper (scientific journal)
The plasmasphere within Earth’s magnetosphere plays a crucial role in space physics, with its electron density distribution being pivotal and strongly influenced by solar activity. Very Low Frequency (VLF) waves, including whistlers, provide valuable insights into this distribution, making the study of their propagation through the plasmasphere essential for predicting space weather impacts on various technologies. In this study, we evaluate the performance of different deep learning model sizes for lightning whistler detection using the YOLO (You Only Look Once) architecture. To achieve this, we transformed the entirety of raw data from the Arase (ERG) Satellite for August 2017 into 2736 images, which were then used to train the models. Our approach involves exposing the models to spectrogram diagrams—visual representations of the frequency content of signals—derived from the Arase Satellite’s WFC (WaveForm Capture) subsystem, with a focus on analyzing whistler-mode plasma waves. We experimented with various model sizes, adjusting epochs, and conducted performance analysis using a partial set of labeled data. The testing phase confirmed the effectiveness of the models, with YOLOv5n emerging as the optimal choice due to its compact size (3.7 MB) and impressive detection speed, making it suitable for resource-constrained applications. Despite challenges such as image quality and the detection of smaller whistlers, YOLOv5n demonstrated commendable accuracy in identifying scenarios with simple shapes, thereby contributing to a deeper understanding of whistlers’ impact on Earth’s magnetosphere and fulfilling the core objectives of this study.
DOI： 10.3390/rs16224264
-
Complex Number Assignment in the Topology Method for Heartbeat Interval Estimation Using Millimeter-Wave Radar Reviewed International journal
Yuji Tanaka, Kimitaka Sumi, Itsuki Iwata, Takuya Sakamoto
IEEE Sensors Letters 8 ( 3 ) 1 - 4 2024.03
Authorship：Lead author, Corresponding author Publishing type：Research paper (scientific journal)
The topology method is an algorithm for accurate estimation of instantaneous
heartbeat intervals using millimeter-wave radar signals. In this model, feature
points are extracted from the skin displacement waveforms generated by
heartbeats and a complex number is assigned to each feature point. However,
these numbers have been assigned empirically and without solid justification.
This study used a simplified model of displacement waveforms to predict the
optimal choice of the complex number assignments to feature points
corresponding to inflection points, and the validity of these numbers was
confirmed using analysis of a publicly available dataset.DOI： 10.1109/LSENS.2024.3364393
Other Link： http://arxiv.org/pdf/2310.11149v1
-
A Bayesian k-vector estimation method for electromagnetic waves in magnetized cold plasma Reviewed International journal
Yuji Tanaka, Mamoru Ota, Shoya Matsuda, Yoshiya Kasahara
URSI Radio Science Bulletin 2021 ( 378 ) 77 - 82 2023.10
Authorship：Lead author, Corresponding author Publishing type：Research paper (scientific journal) Publisher：Institute of Electrical and Electronics Engineers (IEEE)
-
Propagation Analysis Method Considering Angular Spread for Random Electromagnetic Waves in Magnetized Cold Plasma Reviewed
Mamoru Ota, Yuji Tanaka, Yoshiya Kasahara
Radio Science 58 ( 9 ) 2023.09
Publishing type：Research paper (scientific journal) Publisher：American Geophysical Union (AGU)
Abstract
Random waves can be described as the sum of numerous plane waves, and stochastic processes describe their properties. Various methods have been used to widely investigate the propagation characteristics of electromagnetic waves in magnetized cold plasma based on a single‐plane wave approximation. On the other hand, the properties of random waves are difficult to analyze using these methods. Instead, a framework of the wave distribution function (WDF) should be employed. In this study, we provide an explicit expression for an integration kernel in a magnetized cold plasma used for the WDF method. We show that the kernel can be approximated in the case of ultralow frequency/very low frequency (ULF/VLF) parts of whistler‐mode waves with quasi‐parallel propagation. We also propose a method for estimating a WDF representing a directional distribution of the wave energy density based on the principle of maximum entropy using three‐component spectral matrix data of the magnetic field. Based on the insights obtained from the proposed method, we define a quantity called “sharpness,” which provides spreading of wave normal angles. The sharpness is particularly effective in showing the spread of the wave normal angles for random non‐single‐plane waves. Compared with the conventional methods which evaluate the propagation properties (such as planarity), the “sharpness” exhibited a low calculation load and can be implemented easily for onboard processing.DOI： 10.1029/2023rs007673
Other Link： https://agupubs.onlinelibrary.wiley.com/doi/pdf/10.1029/2023RS007673
-
Radar-Based Estimation of Human Body Orientation Using Respiratory Features and Hierarchical Regression Model Reviewed
Wenxu Sun, Shunsuke Iwata, Yuji Tanaka, Takuya Sakamoto
IEEE Sensors Letters 7 ( 9 ) 1 - 4 2023.09
Publishing type：Research paper (scientific journal) Publisher：Institute of Electrical and Electronics Engineers (IEEE)
-
Identification Approach of Arriving Wave Model Based on Likelihood Ratio Test With Different Sensor Noise Levels Reviewed
Yuji Tanaka, Mamoru Ota, Yoshiya Kasahara
Radio Science 57 ( 8 ) 2022.08
Authorship：Lead author, Corresponding author Publishing type：Research paper (scientific journal)
DOI： 10.1029/2022RS007427
-
Noise Integration Kernel Design for the Wave Distribution Function Method: Robust Direction Finding With Different Sensor Noise Levels Reviewed
Yuji Tanaka, Mamoru Ota, Yoshiya Kasahara
Radio Science 56 ( 9 ) 2021.09
Authorship：Lead author Publishing type：Research paper (scientific journal)
DOI： 10.1029/2021RS007291
-
あけぼの衛星の大規模データセットを用いた自然波動の分類に関する研究 Reviewed
J101-D ( 1 ) 225 - 234 2018.01
Authorship：Lead author Publishing type：Research paper (scientific journal)
Misc
-
Individual Identification Using Radar-Measured Respiratory and Heartbeat Features
Haruto Kobayashi, Yuji Tanaka, Takuya Sakamoto
2024.08
This study proposes a method for radar-based identification of individuals
using a combination of their respiratory and heartbeat features. In the
proposed method, the target individual's respiratory features are extracted
using the modified raised-cosine-waveform model and their heartbeat features
are extracted using the mel-frequency cepstral analysis technique. To identify
a suitable combination of features and a classifier, we compare the
performances of nine methods based on various combinations of three feature
vectors with three classifiers. The accuracy of the proposed method in
performing individual identification is evaluated using a 79-GHz
millimeter-wave radar system with an antenna array in two experimental
scenarios and we demonstrate the importance of use of the combination of the
respiratory and heartbeat features in achieving accurate identification of
individuals. The proposed method achieves accuracy of 96.33% when applied to a
five-day dataset of six participants and 99.39% when applied to a public
one-day dataset of thirty participants.Other Link： http://arxiv.org/pdf/2408.00972v1
-
Accurate Radar-Based Heartbeat Measurement Using Higher Harmonic Components
Itsuki Iwata, Kimitaka Sumi, Yuji Tanaka, Takuya Sakamoto
2024.07
This study proposes a radar-based heartbeat measurement method that uses the
absolute value of the second derivative of the complex radar signal, rather
than its phase, and the variational mode extraction method, which is a type of
mode decomposition algorithm. We show that the proposed second-derivative-based
approach can amplify the heartbeat component in radar signals effectively and
also confirm that use of the variational mode extraction method represents an
efficient way to emphasize the heartbeat component amplified via the
second-derivative-based approach. We demonstrate estimation of the heart
interbeat intervals using the proposed approach in combination with the
topology method, which is an accurate interbeat interval estimation method. The
performance of the proposed method is evaluated quantitatively using data
obtained from eleven participants that were measured using a millimeter-wave
radar system. When compared with conventional methods based on the phase of the
complex radar signal, our proposed method can achieve higher accuracy when
estimating the heart interbeat intervals; the correlation coefficient for the
proposed method was increased by 0.20 and the root-mean-square error decreased
by 23%.Other Link： http://arxiv.org/pdf/2407.07380v1
Presentations
-
上肢電気刺激における体性感覚誘発電位に対する波源推定
多田 悠希, 田中 裕士, 小寺 紗千子, 和坂 俊昭, 平田 晃正
2024 年電子情報通信学会ソサイエティ大会 2024.09
-
Individual Identification Using Millimeter-Wave Radar Based on Respiratory Features
Haruto Kobayashi, Yuji Tanaka, Takuya Sakamoto
2024.07
-
Accurate Measurement of Heartbeat by Selecting Radar Echoes Based on Spectral Features
Nobukazu Konishi, Itsuki Iwata, Yuji Tanaka, Takuya Sakamoto
2024.07
-
Accurate Radar-Based Detection of Sleep Apnea Using Overlapping Time-Interval Averaging
Kodai Hasegawa, Yuji Tanaka, Shigeaki Okumura, Hirofumi Taki, Hironobu Sunadome, Satoshi Hamada, Susumu Sato, Kazuo Chin, Takuya Sakamoto
2024.07
-
A Study on Weighting of Amplitude and Phase of Respiratory Displacement in Body Orientation Estimation with Millimeter Wave Radar
Yuji Tanaka, Wenxu Sun, Takuya Sakamoto, Akimasa Hirata
2024.07
-
Initiatives for Physiological Modeling with Millimeter-Wave Radar Invited
Yuji Tanaka, Takuya Sakamoto
2024.06
Event date： 2024.06
Language：Japanese Presentation type：Oral presentation (invited, special)
-
Numerical Evaluation of Source Localization Method Based on Electric Field Analysis Using High-Resolution Human Head Model
Masamune Niitsu, Yuji Tanaka, Chun-Ren Phang, Sachiko Kodera, Toshiaki Wasaka, Akimasa Hirata
2024.04
-
ミリ波レーダによる親子の生体データ可視化の一検討 Invited
田中 裕士, 小林 悠人, 大島 夕侑, 田中 友香里, 明和 政子, 阪本 卓也
日本発達心理学会第 35 回大会 2024.03
Event date： 2024.03
Language：Japanese Presentation type：Symposium, workshop panel (nominated)
-
A Study on Integral Interval in Synthetic Aperture Radar Imaging for Near-Field Finite-Size Targets
Masaya Kato, Yuji Tanaka, Takuya Sakamoto
2024.03
-
Radar-Based People Counting and Association of Multiple Echoes Using Correlation of Physiological Signals
Shoma Nagata, Yuji Tanaka, Takuya, Sakamoto
2024.03
Industrial Property Rights
-
磁気センサ装置、磁気センシング方法およびプログラム
田中 裕士, 高橋 善樹, 高橋 龍平
Applicant：三菱電機株式会社
Application no：JP2020017571 Date applied：2020.04
Patent/Registration no：特許第7094472号 Date registered：2022.06
-
適応制御装置、適応信号処理装置及びアダプティブアレイアンテナシステム
田中 裕士, 高橋 善樹, 高橋 龍平
Applicant：三菱電機株式会社
Application no：JP2019032982 Date applied：2019.08
Publication no：WO2021-038620 Date published：2021.03
-
適応制御装置、適応信号処理装置及びアダプティブアレイアンテナシステム
田中 裕士, 高橋 善樹, 高橋 龍平
Applicant：三菱電機株式会社
Application no：特願2021-541756 Date applied：2019.08
Patent/Registration no：特許第6964833号 Date registered：2021.10
Awards
-
The Second Prize of the URSI-JRSM 2022 Student Paper Competition
2022.09 Japan National Committee of URSI
Yuji Tanaka
-
Student Presentation Award, Joint conference of Hokuriku chapters of Electrical and information Societies
2022.02 The Institute of Electronics, Information and Communication Engineers Hokuriku
Yuji Tanaka
Scientific Research Funds Acquisition Results
-
レーダによる生体情報計測のための特徴点解析手法の確立
Grant number：24K17286 2024.04 - 2027.03
日本学術振興会 科学研究費助成事業 若手研究
田中 裕士
Authorship：Principal investigator
Grant amount：\4550000 （ Direct Cost: \3500000 、 Indirect Cost：\1050000 ）
-
レーダによる人体計測のための体方位モデリング手法の確立
Grant number：23K19119 2023.08 - 2025.03
日本学術振興会 科学研究費助成事業 研究活動スタート支援
田中 裕士
Authorship：Principal investigator
Grant amount：\2860000 （ Direct Cost: \2200000 、 Indirect Cost：\660000 ）
Other External Funds
-
ミリ波レーダによる長期間人体計測のための生体情報分離手法開発
Grant number：24-III5048 2024.11 - 2026.03
公益財団法人 テルモ生命科学振興財団
-
Support for Pioneering Research Initiated by the Next Generation (SPRING)
2021.10 - 2023.03
Teaching Experience
-
Experiments on Introductory Electrical and Electronic Engineering
2024.04 Institution：Nagoya Institute of Technology
Level：Undergraduate (specialized)
-
Advanced Experiments on Electrical and Electronic Engineering
2024.04 Institution：Nagoya Institute of Technology
Level：Undergraduate (specialized)
-
Advanced Practice of Electrical and Electronic Engineering
2023.04 - 2024.03 Institution：Kyoto University
Level：Undergraduate (specialized)
-
Practice of Electrical and Electronic Engineering
2023.04 - 2024.03 Institution：Kyoto University
Level：Undergraduate (specialized)