Di Wu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Di Wu has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Shinichiro Ohnuki
College of Science and Technology, Nihon University, Tokyo, Japan
5 shared publications - Seiya Kishimoto
College of Science and Technology, Nihon University, Tokyo, Japan
3 shared publications T. Yamaguchi
Tokyo Metropolitan Industrial Technology Research Institute,Japan
2 shared publications- Jie Yang
Institute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, Shanghai 200240, China
1 shared publication - Jingxiang Xu
College of Engineering Science and Technology, Shanghai Ocean University, Shanghai 201306, China
1 shared publication - Jun Shibayama
Hosei University
1 shared publication
Latest Publications
- 1 December 2023
- Published by Royal Society of Chemistry (RSC) in Physical Chemistry Chemical Physics
- Vol. 25 (48) , 33094-33103
A performance degradation model of nitrile butadiene rubber (NBR) O-rings for hydrogen permeation effects under high-pressure
The integration of molecular chain changes on a microscopic scale to achieve macroscopic performance is crucial in degradation processes concerning O-ring seals.
- 1 December 2023
- Published by Royal Society of Chemistry (RSC) in Physical Chemistry Chemical Physics
- Vol. 25 (48) , 33094-33103
A performance degradation model of nitrile butadiene rubber (NBR) O-rings for hydrogen permeation effects under high-pressure
The integration of molecular chain changes on a microscopic scale to achieve macroscopic performance is crucial in degradation processes concerning O-ring seals.
- 17 November 2022
- Published by Elsevier in Engineering Failure Analysis
- Vol. 144, 106918
Numerical simulation and verification of goaf morphology evolution and surface subsidence in a mine
- 17 November 2022
- Published by Elsevier in Engineering Failure Analysis
- Vol. 144, 106918
Numerical simulation and verification of goaf morphology evolution and surface subsidence in a mine
- Open Access
Transfer Learning on Electromyography (EMG) Tasks: Approaches and Beyond
Machine learning on electromyography (EMG) has recently achieved remarkable success on a variety of tasks, while such success relies heavily on the assumption that the training and future data must be of the same data distribution. However, this assumption may not hold in many real-world applications. Model calibration is required via data re-collection and label annotation, which is generally very expensive and time-consuming. To address this problem, transfer learning (TL), which aims to improve target learners' performance by transferring the knowledge from related source domains, is emerging as a new paradigm to reduce the amount of calibration effort. In this survey, we assess the eligibility of more than fifty published peer-reviewed representative transfer learning approaches for EMG applications. Unlike previous surveys on purely transfer learning or EMG-based machine learning, this survey aims to provide an insight into the biological foundations of existing transfer learning methods on EMG-related analysis. In specific, we first introduce the physiological structure of the muscles and the EMG generating mechanism, and the recording of EMG to provide biological insights behind existing transfer learning approaches. Further, we categorize existing research endeavors into data based, model based, training scheme based, and adversarial based. This survey systematically summarizes and categorizes existing transfer learning approaches for EMG related machine learning applications. In addition, we discuss possible drawbacks of existing works and point out the future direction of better EMG transfer learning algorithms to enhance practicality for real-world applications.
Open AccessTransfer Learning on Electromyography (EMG) Tasks: Approaches and Beyond
Machine learning on electromyography (EMG) has recently achieved remarkable success on a variety of tasks, while such success relies heavily on the assumption that the training and future data must be of the same data distribution. However, this assumption may not hold in many real-world applications. Model calibration is required via data re-collection and label annotation, which is generally very expensive and time-consuming. To address this problem, transfer learning (TL), which aims to improve target learners' performance by transferring the knowledge from related source domains, is emerging as a new paradigm to reduce the amount of calibration effort. In this survey, we assess the eligibility of more than fifty published peer-reviewed representative transfer learning approaches for EMG applications. Unlike previous surveys on purely transfer learning or EMG-based machine learning, this survey aims to provide an insight into the biological foundations of existing transfer learning methods on EMG-related analysis. In specific, we first introduce the physiological structure of the muscles and the EMG generating mechanism, and the recording of EMG to provide biological insights behind existing transfer learning approaches. Further, we categorize existing research endeavors into data based, model based, training scheme based, and adversarial based. This survey systematically summarizes and categorizes existing transfer learning approaches for EMG related machine learning applications. In addition, we discuss possible drawbacks of existing works and point out the future direction of better EMG transfer learning algorithms to enhance practicality for real-world applications.
- 15 July 2022
- Published by IOP Publishing in Japanese Journal of Applied Physics
- Vol. 61 (8) , 082002
Selectivity of dominant surface plasmon polariton modes using thin metal waveguides with grating structures
The conversion of incident light to surface plasmon polaritons (SPPs) was verified for a dielectric-metal-dielectric waveguide with a grating structure. The dominant SPP mode with respect to the propagation distance could be selected by adjusting the thickness of the gold film. To realize highly efficient light-to-SPPs conversion, the grating parameters for a realistic length of plasmonic waveguides were optimized using a recently developed numerical technique, FDCFD-FILT. This novel technique is the combination of finite-difference complex-frequency-domain (FDCFD) with fast inversion of laplace transform (FILT) and a powerful tool for designing plasmonic waveguides and investigating propagation characteristics efficiently.
- 15 July 2022
- Published by IOP Publishing in Japanese Journal of Applied Physics
- Vol. 61 (8) , 082002
Selectivity of dominant surface plasmon polariton modes using thin metal waveguides with grating structures
The conversion of incident light to surface plasmon polaritons (SPPs) was verified for a dielectric-metal-dielectric waveguide with a grating structure. The dominant SPP mode with respect to the propagation distance could be selected by adjusting the thickness of the gold film. To realize highly efficient light-to-SPPs conversion, the grating parameters for a realistic length of plasmonic waveguides were optimized using a recently developed numerical technique, FDCFD-FILT. This novel technique is the combination of finite-difference complex-frequency-domain (FDCFD) with fast inversion of laplace transform (FILT) and a powerful tool for designing plasmonic waveguides and investigating propagation characteristics efficiently.
- Open Access
- 10 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Multiscale and Multiphysics Computational Techniques
- Vol. 7, 135-141
Error-Controllable Scheme for the LOD-FDTD Method
The implicit locally one-dimensional finite-difference time-domain (LOD-FDTD) method is useful for designing plasmonic devices and waveguide structures. By using a large timestep size, the implicit LOD-FDTD method can reduce the computational time; however, this involves a trade-off with accuracy. To overcome this trade-off, we propose an error-controllable scheme for the LOD-FDTD method, wherein the fast inverse Laplace transform is employed to generate the electromagnetic field in arbitrary time domain from that in complex frequency domain. Compared to the conventional LOD-FDTD method, our scheme provides higher accuracy with more efficient calculations.
- 10 June 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Multiscale and Multiphysics Computational Techniques
- Vol. 7, 135-141
Open AccessError-Controllable Scheme for the LOD-FDTD Method
The implicit locally one-dimensional finite-difference time-domain (LOD-FDTD) method is useful for designing plasmonic devices and waveguide structures. By using a large timestep size, the implicit LOD-FDTD method can reduce the computational time; however, this involves a trade-off with accuracy. To overcome this trade-off, we propose an error-controllable scheme for the LOD-FDTD method, wherein the fast inverse Laplace transform is employed to generate the electromagnetic field in arbitrary time domain from that in complex frequency domain. Compared to the conventional LOD-FDTD method, our scheme provides higher accuracy with more efficient calculations.
- Open Access
- 31 August 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Antennas and Wireless Propagation Letters
- Vol. 19 (12) , 2018-2022
Optimal Parallel Algorithm of Fast Inverse Laplace Transform for Electromagnetic Analysis
Fast Inverse Laplace Transform (FILT) is a promising technique to perform Laplace inverse transform numerically. This paper describes an efficient parallel algorithm of FILT for application in various electromagnetic problems. The proposed algorithm can be easily combined with existing computational methods used in electromagnetics, to realize efficient time and frequency domain analyses.
- 31 August 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Antennas and Wireless Propagation Letters
- Vol. 19 (12) , 2018-2022
Open AccessOptimal Parallel Algorithm of Fast Inverse Laplace Transform for Electromagnetic Analysis
Fast Inverse Laplace Transform (FILT) is a promising technique to perform Laplace inverse transform numerically. This paper describes an efficient parallel algorithm of FILT for application in various electromagnetic problems. The proposed algorithm can be easily combined with existing computational methods used in electromagnetics, to realize efficient time and frequency domain analyses.
- 1 December 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1063-1069
Time-division Efficient Parallel Algorithm for Designing Metallic Slabs for Quantum Walk
Parallel computing techniques play an important role in computational electromagnetics for performing fast and efficient simulation. In conventional techniques, the entire space is divided into many smaller parts whose computational loads are distributed into cluster. This paper reports a novel parallel technique, i.e., a time-division algorithm. In our method, EM field in frequency-domain is obtained by the finite-difference complex-frequency-domain (FDCFD) method. The field in the complex-frequency-domain is transformed into the time domain by fast inverse Laplace transform (FILT). The advantage of the proposed technique is that we can efficiently obtain the response at required observation time by parallel computing. As a numerical example of applications, a plasmonic waveguide array (PWA) using metallic slabs is designed for quantum walk toward developing quantum information technology. Properties of PWA will be discussed in comparison with classical random walk.
- 1 December 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1063-1069
Time-division Efficient Parallel Algorithm for Designing Metallic Slabs for Quantum Walk
Parallel computing techniques play an important role in computational electromagnetics for performing fast and efficient simulation. In conventional techniques, the entire space is divided into many smaller parts whose computational loads are distributed into cluster. This paper reports a novel parallel technique, i.e., a time-division algorithm. In our method, EM field in frequency-domain is obtained by the finite-difference complex-frequency-domain (FDCFD) method. The field in the complex-frequency-domain is transformed into the time domain by fast inverse Laplace transform (FILT). The advantage of the proposed technique is that we can efficiently obtain the response at required observation time by parallel computing. As a numerical example of applications, a plasmonic waveguide array (PWA) using metallic slabs is designed for quantum walk toward developing quantum information technology. Properties of PWA will be discussed in comparison with classical random walk.
- 18 April 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Technology Letters
- Vol. 30 (11) , 1024-1027
Finite-Difference Complex-Frequency-Domain Method for Optical and Plasmonic Analyses
The finite-difference frequency-domain (FDFD) method is one of the most effective methods for obtaining the frequency responses of specific optical components, such as thin films and guided wave structures. This paper describes an efficient algorithm based on the FDFD method to perform optical and plasmonic analyses. Solutions to the Maxwell’s equations in the complex-frequency-domain are obtained by the finite-difference scheme and are numerically transformed into the time domain using the fast inverse Laplace transform. Reliable time and frequency responses of electromagnetic waves can be obtained through a simple and fast procedure. Furthermore, the time increment and observation time can be selected freely, while maintaining the accuracy. These are verified by performing the electric near-field analyses of a metallic cylinder. The computational results by our method are compared with the exact solution and the FDTD method. Some details of performing inverse Laplace transform also discussed.
- 18 April 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Technology Letters
- Vol. 30 (11) , 1024-1027
Finite-Difference Complex-Frequency-Domain Method for Optical and Plasmonic Analyses
The finite-difference frequency-domain (FDFD) method is one of the most effective methods for obtaining the frequency responses of specific optical components, such as thin films and guided wave structures. This paper describes an efficient algorithm based on the FDFD method to perform optical and plasmonic analyses. Solutions to the Maxwell’s equations in the complex-frequency-domain are obtained by the finite-difference scheme and are numerically transformed into the time domain using the fast inverse Laplace transform. Reliable time and frequency responses of electromagnetic waves can be obtained through a simple and fast procedure. Furthermore, the time increment and observation time can be selected freely, while maintaining the accuracy. These are verified by performing the electric near-field analyses of a metallic cylinder. The computational results by our method are compared with the exact solution and the FDTD method. Some details of performing inverse Laplace transform also discussed.