State estimation of learner in a learning situation where gaze target changes

IEICE Technical Report; IEICE Tech. Rep.

Published On 2020/1/18

(in English) Support from teachers and assistants is essential for children with learning disabilities. It is not easy for teachers to provide appropriate support to a plurality of students in classes in a general classroom. Therefore, we have proposed a system for estimating the degree of concentration based on the matching between the learner's gaze and the gaze target, and a learning status notification system based on the estimation. In this study, we developed a system that can switch the gaze target using gestures measured based on the acceleration sensor worn on the wrist. We show the experimental results on the estimation of the degree of concentration when the gaze target changes

Journal

IEICE Technical Report; IEICE Tech. Rep.

Published On

2020/1/18

Volume

119

Issue

394

Page

23-27

Authors

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Position

H-Index(all)

44

H-Index(since 2020)

19

I-10 Index(all)

0

I-10 Index(since 2020)

0

Citation(all)

0

Citation(since 2020)

0

Cited By

0

Research Interests

Odor sensor

Dosimeter

Thin films

University Profile Page

Other Articles from authors

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Sensors and Materials

Radio-photoluminescence Properties of Eu-doped CaBPO5

Eu-doped CaBPO5 was synthesized by the solid-state reaction route, and then its radiophotoluminescence properties were studied. On the basis of steady-state and time-resolved photoluminescence (PL) studies, it was confirmed that Eu2+ is formed by X-ray irradiation due to the valence change of the Eu ion (Eu3+→ Eu2+). In addition, for dosimetric applications, it was demonstrated that the PL intensity of Eu2+ is proportional to the accumulated radiation dose, is stable after irradiation, and can be reversed by heat treatment at 500℃ for 100 s.

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Japanese Journal of Applied Physics

Radiophotoluminescence properties of Eu-doped Li2SrSiO4 for radiation detectors

To search for new radiophotoluminescence (RPL) materials, Eu-doped Li 2 SrSiO 4 was synthesized by the solid-state reaction and then studied its RPL properties for radiation detector applications. All the samples were identified to be Li 2 SrSiO 4 by X-ray diffraction. The as-prepared sample shows photoluminescence (PL) due to Eu 3+ while an additional PL band appears due to a formation of Eu 2+ after X-ray irradiation. This indicates that the Eu-doped Li 2 SrSiO 4 has RPL properties. The sensitivity of RPL (or PL intensity of Eu 2+) to X-ray dose is the highest when the Eu concentration is 0.1%, and the dynamic range is 5–10 000 mGy. In addition, the generated Eu 2+ is so stable that only 5% of the signal degrades over 20 min after irradiation. The RPL signal notably decreases to about 30% of the initial value by heating at 500 C for 100 s, and it is reproducible multiple times even after the heat treatment.

2023/12/20

Article Details
Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Journal of Materials Science: Materials in Electronics

ESR studies to elucidate the origin of the RPL emission of undoped-CaSO4

Recent research has shown that CaSO4, without any intentional addition of impurities, exhibits remarkable radiophotoluminescence (RPL) properties. However, the origin of the RPL in CaSO4 has yet to be fully understood. ESR measurements were carried out to identify the defect species responsible for RPL and elucidate the luminescence’s origin. ESR measurements on undoped-CaSO4 with different radiation sensitivities showed SO4−, SO2−+SO3−, PO42−, Si=O3 in P2O5–SiO2 (E’ center), Al2O3 hole center and Mn2+ defect species in CaSO4. This means that CaSO4 contains Al2O3, Mn2+, P, and Si as impurities, which are not intentionally added to CaSO4. As the results of this study, the different impurity states may influence the radiation sensitivity of the materials. It is suggested that the deliberate addition of P and Si elements, Al2O3 and Mn2+ could control the RPL properties.

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Materials Research Bulletin

X-ray induced reduction of Sm3+ into Sm2+ in Li2CaSiO4 and its potential for radiation measurement applications

In Sm3+-doped Li2CaSiO4, the Sm3+ ion is reduced to the Sm2+ state by X-ray irradiation. The valence change was confirmed by the difference in their photoluminescence spectra, measured before and after irradiation. Only luminescence peaks due to the 4f-4f transitions of Sm3+ are observed before the irradiation while an additional luminescence band is observed at 900 nm due to the 5d-4f transition of Sm2+. The reaction is that a fraction of X-ray generated electrons is captured by Sm3+ ions, which convert to the Sm2+ state. The generated Sm2+ state is fairly stable while it is effectively reversed back to the Sm3+ state by a heat treatment at an elevated temperature. A numerical approximation revealed the activation energy to be 0.40 and 0.52 eV. Even after the heat-treatment, the reduction of Sm3+ to Sm2+ is reproducible multiple times, which shows good potential for using radiation measurement applications.

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Journal of Materials Science: Materials in Electronics

Radio-photoluminescence properties of Eu-doped SrAl2O4 ceramics and thermally assisted charge transfer process

Radio-photoluminescence (RPL) properties of Eu-doped SrAl2O4 are studied focusing on the generation of a luminescent center by ionizing radiation of X-ray and its extinction by a thermally assisted charge transfer. The single phases of SrAl2O4 doped with Eu at concentrations of 0.05–1.0% are synthesized via solid-state reactions, as confirmed by X-ray diffractions analyses. They show photoluminescence (PL) due to both Eu3+ and Eu2+, and the PL intensity due to Eu3+ decreases after X-ray irradiation while the one due to Eu2+ notably increases. This indicates that Eu2+ as a luminescence center is generated as RPL through reduction from Eu3+ to Eu2+ by ionizing radiation of X-ray. Particularly, the 0.1% Eu-doped sample shows the highest sensitivity to X-ray dose, and the lowest detection limit is estimated to be 0.3 Gy with the present reader system. The RPL signal is reasonably stable, and it depends on …

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Recent advances in radiophotoluminescence materials for luminescence dosimetry

Radiophotoluminescence (RPL) is a phenomenon in which a luminescence centre is formed in a medium upon interaction with ionizing radiation. The RPL is observed by a conventional photoluminescence (PL) technique, and the PL intensity is proportional to the dose. With the latter feature, the RPL has found successful applications in personnel and environmental dosimetry. However, the conventional materials considered for radiation measurements may be limited to Ag-doped phosphate glass, LiF, and Al 2 O 3: C, Mg. The recent research works, however, have found a number of additional RPL materials for luminescence dosimetry. The aim of the present paper is to review the series of newly reported RPL materials and potential applications in dosimetry.

2022/11/22

Article Details
Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Optically stimulated luminescence dosimeters: principles, phosphors and applications

This review paper discusses the basics of the optically stimulated luminescence (OSL) phenomenon, phosphors for OSL, functional design of OSL phosphors, and application areas of the OSL phenomenon, such as radiation dose monitoring for individual and environmental, computed radiography for medical and dental diagnosis, dating, security, archaeological dating, and autoradiography for drug discovery. Finally, the future prospects of the OSL phenomenon as passive-type dosimeters are mentioned.

2022/11/21

Article Details
Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

IEEJ Transactions on Sensors and Micromachines

Proposal of Temperature Behavior Analysis and Temperature Dependence Reduction of Stem Water Content Variation Sensor

Proposal of Temperature Behavior Analysis and Temperature Dependence Reduction of Stem Water Content Variation Sensor - NASA/ADS Now on home page ads icon ads Enable full ADS view NASA/ADS Proposal of Temperature Behavior Analysis and Temperature Dependence Reduction of Stem Water Content Variation Sensor Ueno, Chiemi ; Hirasawa, Kazuki ; Takei, Yoshinori ; Nanto, Hidehito Abstract Publication: IEEJ Transactions on Sensors and Micromachines Pub Date: June 2022 DOI: 10.1541/ieejsmas.142.117 Bibcode: 2022IJTSM.142..117U No Sources Found © The SAO/NASA Astrophysics Data System adshelp[at]cfa.harvard.edu The ADS is operated by the Smithsonian Astrophysical Observatory under NASA Cooperative Agreement NNX16AC86A NASA logo Smithsonian logo Resources About ADS ADS Help What's New Careers@ADS Social @adsabs ADS Blog Project Switch to full ADS Is …

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Sensors and Materials

Radiophotoluminescence Imaging Reader for Passive Dosimetry

In this study, we constructed an image reader system for two-dimensional radiation dosimetry based on radiophotoluminescence (RPL). The reader system mainly consists of a Peltier-cooled CMOS camera and an LED excitation source, as well as control units, and it offers an arbitrary excitation wavelength (eg, 365, 405, 460, 530, and 630 nm) and a wide spectral detection range (200–1000 nm) to be used for a wide range of prototype image detectors having different spectral features. Using the developed system, we were able to successfully obtain X-ray projection images recorded on a commercial RPL detector [Ag-doped phosphate glass (Ag-PG) plate] as well as custom-made prototypes of flexible imaging plates (IPs). Furthermore, the reader system is applicable not only for RPL imaging but also for conventional optically stimulated luminescence (OSL) imaging, which was demonstrated by using a …

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Japanese Journal of Applied Physics

RPL properties of samarium-doped CaSO4

Radiophotoluminescence (RPL) properties of Sm-doped CaSO 4 for radiation dosimetry applications are reported. The samples with varying Sm concentrations are prepared via the solid-state reaction process. The as-prepared samples show photoluminescence due to typical 4f–f transitions of Sm 3+ whereas, after X-ray irradiation, additional emission features appear with a broad band peaking at 630 nm as well as a set of multiple sharp lines across 680–820 nm, which are attributed to the 5d–4f and 4f–4f transitions of Sm 2+, respectively. Therefore, the RPL in the present material system relies on the generation of Sm 2+ centers. The sensitivity is about 3 times lower than that of Ag-doped phosphate glass, but no fading and build-up of signal are evident even immediately after the irradiation. The signal is reversible by heat-treatment at 500 C, and is reproducible even after the thermal erasure, especially when …

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Phosphors for Radiation Detectors

radiophotoluminescence (RPL)

This chapter covers a wide range of radio‐photoluminescence (RPL) technologies, from fundamentals to applications. It presents the definition of the term of RPL and distinguishes it from thermally‐stimulated luminescence and optically‐stimulated luminescence as a different type of dosimetric technique using a phosphor medium. RPL is a very powerful tool when the measurement environment requires exposure to light or heat. The chapter introduces a list of RPL materials and their fundamental properties, as reported in the literature. The RPL center is typically very stable and the resulting luminescence is due to electronic transitions within the RPL center. The chapter also introduces the principal properties as well as applications of four selected RPL materials with many illustrative materials. These include: Ag‐doped sodium‐aluminophosphate glasses, Al …

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Phosphors for Radiation Detectors

Optically‐Stimulated Luminescent Dosimeters

The luminescence phenomenon due to the thermal stimulation is called thermally‐stimulated luminescence (TSL), while the one due to the optical stimulation is called optically‐stimulated luminescence (OSL). This chapter focuses on the principles and applications of OSL, which are utilized today in personnel dose monitoring, environmental dose monitoring, and geographical dating. OSL phenomenon has been used in personnel dose monitoring and imaging plates for computed radiography since the 1980s. A dosimeter using C‐doped Al 2 O 3 (Al 2 O 3 :C) developed by Landauer, Inc. is one of those passive‐type dosimeters using OSL phenomenon for personnel dose monitoring applications. Al 2 O 3 :C …

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Journal of The Institute of Electrical Engineers of Japan

What Did the Great Persons Who Changed the World Think and Say?

What Did the Great Persons Who Changed the World Think and Say? - NASA/ADS Now on home page ads icon ads Enable full ADS view NASA/ADS What Did the Great Persons Who Changed the World Think and Say? Nanto, Hidehito Abstract Publication: Journal of The Institute of Electrical Engineers of Japan Pub Date: January 2022 DOI: 10.1541/ieejjournal.142.34 Bibcode: 2022JIEEJ.142...34N full text sources Publisher | © The SAO/NASA Astrophysics Data System adshelp[at]cfa.harvard.edu The ADS is operated by the Smithsonian Astrophysical Observatory under NASA Cooperative Agreement NNX16AC86A NASA logo Smithsonian logo Resources About ADS ADS Help What's New Careers@ADS Social @adsabs ADS Blog Project Switch to full ADS Is ADS down? (or is it just me...) Smithsonian Institution Smithsonian Privacy Notice Smithsonian Terms of Use Smithsonian Astrophysical Observatory …

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Physical Review Research

Catalog of gamma-ray glows during four winter seasons in Japan

In 2015, the Gamma-Ray Observation of Winter Thunderstorms (GROWTH) collaboration launched a mapping observation campaign for high-energy atmospheric phenomena related to thunderstorms and lightning discharges. This campaign has developed a detection network of gamma rays with up to 10 radiation monitors installed in the cities of Kanazawa and Komatsu, Ishikawa Prefecture, Japan, where low-charge-center winter thunderstorms frequently occur. During four winter seasons from October 2016 to April 2020, a total of 70 gamma-ray glows, ie, minute-lasting bursts of gamma rays originating from thunderclouds, were detected. Their average duration is 58.9 s. Among the detected events, 77% were observed at night. The gamma-ray glows can be classified into temporally symmetric, temporally asymmetric, and lightning-terminated types based on their count-rate histories. An averaged energy …

2021/11/15

Article Details
Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Materials Letters

Radiation-induced reduction of Eu3+ doped in SrAl2O4

In this letter, we report radio-photoluminescence (RPL) properties of Eu-doped SrAl2O4. The RPL is a phenomenon in which a new luminescence center is generated by ionizing radiation. The material was synthesized via the solid-state reaction. The synthesized sample shows photoluminescence (PL) due to the 4f-4f transition of Eu3+ and the 4f65d1 → 4f7 transition of Eu2+. After X-ray irradiation, the former emission intensity decreases while the latter one increases, and the extent of the change depends on the radiation dose. The change in the RPL is attributed to a reduction of Eu3+ by trapping an X-ray generated electron (Eu3+→Eu2+). In addition, a heat-treatment effectively reverses the above reaction by detrapping an electron from the Eu2+ center.

2021/11/15

Article Details
Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Sensors & Materials

TSL/OSL/RPL Automated and Integrated Measurement System (TORAIMS).

Thermally stimulated luminescence (TSL), optically stimulated luminescence (OSL), and radio-photoluminescence (RPL) are radiation-induced luminescence phenomena, and they are widely used for radiation dosimetry today. Despite their successful applications, the R&D of conventional and new phosphor materials is still actively performed in a wide range of research fields from basic science to application. Although these phenomena are considered to be different, they share common physical processes such as the generation, trapping, and transfer of electronic charges. In order to have a deeper look into these phenomena, it is important to comprehensively characterize those of a single material. For this, we have constructed a prototype TSL/OSL/RPL automated and integrated measurement system (TORAIMS), and then the system performance has been studied. The system offers automated and integrated …

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Radioprotection

Characterization of optically-stimulated luminescence properties by NaCl: Eu2+ crystal and the thermal response

[en] Landauer optically stimulated luminescence (OSL) technology nanoDot dosimeters (OSLDs) are characterized for use in mammography screening at various tube voltages, mAs values and target/filter combinations. The average glandular dose (AGD) for a 50-mm breast, based on the representative compressed breast thickness of a 45-mm polymethyl methacrylate (PMMA) phantom, is assessed using OSLDs with different beam conditions. Further, the linearity of the OSLD response is measured and angular dependence tests are performed for various tube potentials, mAs and target/filter combinations. The breast-absorbed doses are measured at various depths for a 32-kVp X-ray beam at 100 mAs, with a Mo/Rh target/filter combination. The measured incident air kerma values at different lateral positions exhibit a maximum deviation of 6%, and the average relative response of the OSLDs at the reference point (center) with respect to various lateral positions is found to be 1.001±0.09%. The calculated AGD values are in the 1.3±0.1-3.5±0.2 mGy range, depending on the tube potential, tube loading and target/filter combinations. An exposure setup featuring the auto-exposure control (AEC) mode, 28 kVp, 73.8 mAs, and a Mo/Rh target/filter combination may be preferred for mammography screening for a compressed breast thickness of 45 mm.(authors)

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

電気学会研究会資料

Transient response analysis and feature extraction of semiconductor gas sensors for an application to machine olfaction

瀬戸就一 他: 半導体式ガスセンサの過渡応答出力の解析, 電気学会論文誌 E センサマイクロマシン準部門誌 Vol. 125 No. 3, p. 129-134 (2005)

Hidehito Nanto

Hidehito Nanto

Kanazawa Institute of Technology

Radiation Measurements

Physical and dosimetric characteristics of radiophotoluminescent glass from two-photon excitation microscopy

We investigate the physical and dosimetric characteristics of Ag-activated phosphate glass exposed to proton beams with various energies. We present time-resolved blue and orange radiophotoluminescent (RPL) spectra and examine the dependence of the peak wavelengths and RPL area intensities on gate time. Moreover, we consider various micro-scale 3D patterns written on Ag-activated glass by a focused proton beam and evaluate their orange and blue RPL axial-profiles using two-photon excitation microscopy. Finally, we discuss the remarkable broadening of the Bragg curve observed in Ag-activated glass and argue that it should be attributed to the saturation effect and/or non-linearity effect from the viewpoint of the relation between the dose delivered and the signal.

Other articles from IEICE Technical Report; IEICE Tech. Rep. journal

Jin Nakazato

Jin Nakazato

Tokyo Institute of Technology

IEICE Technical Report; IEICE Tech. Rep.

User Scheduling based on Expanded Null-Space for Massive MIMO

(in English) This paper proposes a user scheduling scheme for multi-user massive MIMO (MU-mMIMO), particularly effective in time-varying channel environments. MU-MIMO enables spectral efficiency enhancement by multiplexing a number of user terminals (UTs) in the spatial domain.Due to the finite number of base station (BS) antennas, the simultaneous transmission capacity is constrained in MU-MIMO. Therefore, BS must divide UTs into distinct subsets and transmit signals for each subset. When UTs move at high speed, it causes inter-user interference (IUI) and subsequently reduces the spatial multiplexing capability.

Naoya Isoyama

Naoya Isoyama

Nara Institute of Science and Technology

IEICE Technical Report; IEICE Tech. Rep.

Influence of Temperature Stimulus to the Sitting Person's Seat Surface on the Center of Gravity

(in English) In this study, we focus on the deterioration of sitting posture due to VDT work that is spreading in modern society and the associated health problems, and investigate the possibility of improving sitting posture by means of thermal stimulation. Studies have been conducted on physical devices and information presentation methods to improve sitting posture. Physical devices promote effective posture improvement, but the physical deformation limits the user's degree of freedom and may cause stress. On the other hand, information presentation methods have the advantage of easily presenting information, but require the user to pay attention to the information presented, and spontaneous behavior change based on the presented information is not guaranteed to be effective depending on the will of the individual and the situation. In this study, we examine the effects of thermal stimuli on the actual and …

Yuki Matsuda (松田 裕貴)

Yuki Matsuda (松田 裕貴)

Nara Institute of Science and Technology

IEICE Technical Report; IEICE Tech. Rep.

Satisfaction Evaluation of Recommended Intake Calculation Method for Postprandial Hyperglycemia Prevention

(in English) Continuous self-management of environmental factors, including dietary habits to prevent postprandial hyperglycemia is essential. Previously, we proposed a framework for dietary behavior change that involves real-time monitoring of blood glucose levels, predicting postprandial hyperglycemia from planned meals, calculating foods that should be reduced in intake, and presenting the results to users. However, the optimal calculation method for selecting foods and quantities to induce behavioral change while maintaining meal satisfaction for preventing postprandial hyperglycemia remains unclear. To realize the proposed framework, it is necessary to establish calculation methods for recommended intake of each food and evaluate the satisfaction of meals when applying these recommended intake amounts.In this paper, we set multiple recommended intake calculation methods and conducted …

Koichi Kobayashi

Koichi Kobayashi

Hokkaido University

IEICE Technical Report; IEICE Tech. Rep.

Design of Block-Sparse Controllers for Cyber-Physical Systems

(in English) In sensor and actuator networks, it is important to consider not only the choice of sensors/actuators but also that of communication channels. In this paper, we propose an LMI-based design method for structured block-sparse output feedback controllers. First, the sparse reconstruction problems for vectors and matrices are summarized. Next, the design problem is formulated, and an LMI-based solution method is proposed. Finally, a simple example is presented to show the effectiveness of the proposed method.

Mohd Adib Sarijari

Mohd Adib Sarijari

Universiti Teknologi Malaysia

IEICE Technical Report; IEICE Tech. Rep.

An Evaluation of CNN Using Deep Residual Learning for Modulation, 5G, LTE, and WLAN System Classification

(in English) In this study, we investigate and present a deep residual learning for modulation classification. The simulation results show the degradation problem that was exposed due to an increase in network depth and the saturation of accuracy in the modified conventional CNN; however, the proposed CNN has no such degradation. Therefore, the processing burden of the conventional CNN is much larger than the proposed CNN. In the simulation results, the proposed CNN framework achieves almost the same modulation classification accuracy as the normal CNN framework when reducing the processing burden in the proposed one. The better simulation results are shown by adjustment of the parameters using the proposed method in the case of OFDM and single carrier modulation types.

Ching-Ming Lai, IET Fellow/AIE Fellow/IEEE Senior Member

Ching-Ming Lai, IET Fellow/AIE Fellow/IEEE Senior Member

National Chung Hsing University

IEICE Technical Report; IEICE Tech. Rep.

Over 98% Efficiency Floating Four-Phase Interleaved Charge-Pump Bidirectional DC-DC Converter with Wide Buck/Boost Voltage Ratio

(in English) A novel floating four-phase interleaved charge-pump bidirectional dc-dc converter (4F-ICPBDC) with wide Buck/Boost voltage ratio is proposed in this article. The interleaved structure is adopted to reduce the ripple of the current through the low-voltage side, and the floating structure is adopted to achieve the high Buck/Boost voltage conversion ratio. In addition, to achieve balance of the average inductor current within the full duty cycle range, an asymmetric duty limit control (DLC) strategy has been implemented. Furthermore, the operating principle of the converter is analyzed. Finally, a 1 kW prototype is developed with SiC-MOSFETs, which verifies the wide Buck/Boost voltage ratio of this converter between the constant low-voltage side (72 V) and the variable high-voltage side (400-800 V). The maximum efficiencies of the converter are recorded as 98.3% in the Buck mode and 97.6% in the Boost …

Kenta Ninomiya

Kenta Ninomiya

Kyushu University

IEICE Technical Report; IEICE Tech. Rep.

Prediction of Cancer Relapse for Non-Small Cell Lung Cancer Patients Using Clinical and Imaging Features Before Stereotactic Ablative Radiotherapy

(in English) The aim of this study is to predict locoregional relapse and distant metastasis in stage I non-small cell lung cancer patients using clinical and image features before radiotherapy. Image features were obtained from Betti number maps and wavelet decomposition images calculated from CT images. The predictive performances were evaluated by significant differences of the Kaplan-Meier curves for high-risk and low-risk patient groups predicted from the clinical and image features. The clinical and image combined features could have a potential to improve the relapse risk classification.

Keiichi Yasumoto

Keiichi Yasumoto

Nara Institute of Science and Technology

IEICE Technical Report; IEICE Tech. Rep.

A PoI Localization Method using Time-series SNS Posts with Photos

(in English) In recent years, with the easing of COVID-19 movement restrictions, there has been a trend towards the recovery of tourism demand. In tourist destinations, tourists increasingly use social media not only for disseminating information but also for gathering information. It is common for them to decide their destinations based on information posted by others. However, posts with location information typically involve tagging posts with locations registered in existing databases, which means unregistered locations cannot be tagged. This poses a challenge as there is also a trend in social media towards hiding location information for privacy protection. To apply these posts to tourist information services, it is necessary to estimate the Point of Interest (PoI) locations. This study proposes a method for estimating PoI locations based on the temporal sequence of photo and text posts on social media. Experiments …

Fumio Teraoka

Fumio Teraoka

Keio University

IEICE Technical Report; IEICE Tech. Rep.

Basic Evaluation of MECM2M, A M2M System Infrastructure Considering Mobile Nodes as Computing Servers

(in English) We are developing MECM2M, an M2M (Machine-to-Machine) system infrastructure suitable for MEC (Multi-access Edge Computing) environments, considering mobile nodes with both sensor/actuator and compute server functions, such as automobiles and smart phones.Based on the design concept of the Internet, MECM2M provides M2M applications with only the basic functions of acquiring sensor values and instructing actuators to move by specifying areas or nodes.

Shinpei Ogata

Shinpei Ogata

Shinshu University

IEICE Technical Report; IEICE Tech. Rep.

A Method to Create Scenario by Large Language Model for Input Amount Simulation

(in English) One of the means to reasonably reduce the burden of user input to the UI (User Interface) is to introduce usability-enhancing functions (usability functions), but it is difficult to find an appropriate configuration method for them. In order to improve this problem, we have proposed a simulation method to estimate the user input amount from screen transition models and scenarios for a usability function called an input storage function. However, there was a problem that the burden of manual scenario creation by the method user was large. To solve this problem, we propose a method to support scenario creation using a large language model. To evaluate the effectiveness of the proposed method, we attempted to create scenarios using the proposed method for three types of scenarios created in a previous experiment. The results showed that the proposed method was more effective than the methods used in …

Keiichi Yasumoto

Keiichi Yasumoto

Nara Institute of Science and Technology

IEICE Technical Report; IEICE Tech. Rep.

Satisfaction Evaluation of Recommended Intake Calculation Method for Postprandial Hyperglycemia Prevention

(in English) Continuous self-management of environmental factors, including dietary habits to prevent postprandial hyperglycemia is essential. Previously, we proposed a framework for dietary behavior change that involves real-time monitoring of blood glucose levels, predicting postprandial hyperglycemia from planned meals, calculating foods that should be reduced in intake, and presenting the results to users. However, the optimal calculation method for selecting foods and quantities to induce behavioral change while maintaining meal satisfaction for preventing postprandial hyperglycemia remains unclear. To realize the proposed framework, it is necessary to establish calculation methods for recommended intake of each food and evaluate the satisfaction of meals when applying these recommended intake amounts.In this paper, we set multiple recommended intake calculation methods and conducted …

ROZEHA BT. A. RASHID

ROZEHA BT. A. RASHID

Universiti Teknologi Malaysia

IEICE Technical Report; IEICE Tech. Rep.

An Evaluation of CNN Using Deep Residual Learning for Modulation, 5G, LTE, and WLAN System Classification

(in English) In this study, we investigate and present a deep residual learning for modulation classification. The simulation results show the degradation problem that was exposed due to an increase in network depth and the saturation of accuracy in the modified conventional CNN; however, the proposed CNN has no such degradation. Therefore, the processing burden of the conventional CNN is much larger than the proposed CNN. In the simulation results, the proposed CNN framework achieves almost the same modulation classification accuracy as the normal CNN framework when reducing the processing burden in the proposed one. The better simulation results are shown by adjustment of the parameters using the proposed method in the case of OFDM and single carrier modulation types.

Yuki Matsuda (松田 裕貴)

Yuki Matsuda (松田 裕貴)

Nara Institute of Science and Technology

IEICE Technical Report; IEICE Tech. Rep.

A PoI Localization Method using Time-series SNS Posts with Photos

(in English) In recent years, with the easing of COVID-19 movement restrictions, there has been a trend towards the recovery of tourism demand. In tourist destinations, tourists increasingly use social media not only for disseminating information but also for gathering information. It is common for them to decide their destinations based on information posted by others. However, posts with location information typically involve tagging posts with locations registered in existing databases, which means unregistered locations cannot be tagged. This poses a challenge as there is also a trend in social media towards hiding location information for privacy protection. To apply these posts to tourist information services, it is necessary to estimate the Point of Interest (PoI) locations. This study proposes a method for estimating PoI locations based on the temporal sequence of photo and text posts on social media. Experiments …

Akira Uchiyama

Akira Uchiyama

Osaka University

IEICE Technical Report; IEICE Tech. Rep.

A Preliminary Study on Parameter Optimization Using a Backpropagation Algorithm for a Neonatal Thermal Model

(in English) Neonates need temperature management in incubators due to their underdeveloped thermoregulatory functions. Traditional methods using skin-attached probes are challenging due to the neonates' delicate skin. Therefore, efforts are being made to realize non-contact temperature measurement using thermography. However, thermography can be hindered during medical procedures by obstructions like medical staff's hands. To address this, we utilize a human thermal model to simulate body temperature changes, integrating real-time sensor data, including thermography and incubator conditions, to estimate surface and core body temperatures. Our method employs backpropagation for rapid parameter optimization in the human thermal model. The effectiveness was confirmed using actual neonatal data, achieving core body temperature estimation with an average absolute error of 0.032$ degree …

Sinan Chen

Sinan Chen

Kobe University

IEICE Technical Report; IEICE Tech. Rep.

A Study of Promotion Method for Energy-Saving Behavior in Homes with Personalized Adaptive Interaction

(in English) In recent years, climate change, including global warming, has become a serious issue, and Japan is aiming to realize a zero carbon society. Zero carbon means reducing the net emissions of greenhouse gases to zero by balancing the amount emitted and absorbed, which is crucial for reducing CO2 emissions. This study focuses on energy reduction through in-home energy-saving actions, addressing issues such as the lack of acquisition of appliance-specific electricity consumption data, lack of personal adaptation, and oversight of notifications. To address these, a method that integrates a power consumption management service using IoT devices and baselines with a virtual agent (VA) was proposed and implemented. The evaluation experiment confirmed that the introduction of VA led to a reduction in power consumption, and an improvement in energy-saving consciousness and behavioral …

RUBITA SUDIRMAN

RUBITA SUDIRMAN

Universiti Teknologi Malaysia

IEICE Technical Report; IEICE Tech. Rep.

An Evaluation of CNN Using Deep Residual Learning for Modulation, 5G, LTE, and WLAN System Classification

(in English) In this study, we investigate and present a deep residual learning for modulation classification. The simulation results show the degradation problem that was exposed due to an increase in network depth and the saturation of accuracy in the modified conventional CNN; however, the proposed CNN has no such degradation. Therefore, the processing burden of the conventional CNN is much larger than the proposed CNN. In the simulation results, the proposed CNN framework achieves almost the same modulation classification accuracy as the normal CNN framework when reducing the processing burden in the proposed one. The better simulation results are shown by adjustment of the parameters using the proposed method in the case of OFDM and single carrier modulation types.

Shinji Watanabe

Shinji Watanabe

Carnegie Mellon University

IEICE Technical Report; IEICE Tech. Rep.

Evaluating speech generation based on objective measures for text generation

(in English) In the evaluation of speech generation, while subjective judgments have long been the gold standard, objective metrics such as Mel Cepstral Distortion (MCD) and Mean Opinion Score (MOS) prediction models have also been used. These objective metrics are valued for their lower time and financial costs and the ability to compare different results, driving the demand for metrics that correlate well with human subjective judgments. This paper proposes an automatic evaluation method for speech generation based on text generation metrics. Our proposed SpeechBERTScore computes BERTScores on self-supervised speech feature sequences derived from both generated and reference speech. In addition, SpeechBLEU and SpeechTokenDistance define metrics using self-supervised discrete speech tokens. Experimental evaluations of synthesized speech show that our SpeechBERTScore correlates …

Yohei Murakami

Yohei Murakami

Ritsumeikan University

IEICE Technical Report; IEICE Tech. Rep.

Cooperative Agents for Federated Learning of Neural Machine Translation

(in English) Accurate neural machine translation requires large amounts of high-quality bilingual data, however due to the copyright and confidentiality issues, it is difficult to share the bilingual data between different organizations. With federal learning, multiple clients can collaboratively build a neural machine translation model and share only the translation model while keeping their own data confidential. However, if the domain of bilingual data differs among clients, the translation accuracy is not necessarily improved by integrating all translation models. Therefore, we propose a cooperative agent that dynamically selects a cooperative partner and integrates translation models in each aggregation process in cooperative learning. Compared to conventional cooperative learning methods, our method can improve the translation accuracy of each client by 22.7% on average.

TUTOMU MURASE

TUTOMU MURASE

Nagoya University

IEICE Technical Report; IEICE Tech. Rep.

Segmented file delivery method by mobile APs using a combination of overhearing and transmission rate control

(in English) In the event of a large-scale disaster, there is a possibility that the communication infrastructure may not work properly. Therefore, a method of distributing evacuation information that does not depend on the communication infrastructure is required. Ad hoc networks and DTNs are available as communication methods that do not depend on the infrastructure in the event of a disaster. In addition, other research has proposed a method that enable information distribution to terminals over a wide area by using mobile vehicles equipped with wireless LAN access points. However, conventional researches consider information which size is a few Byte. Therefore, they did not consider a large amount of information which size is about 10Mbyte such as the evacuation map to deliver for multiple terminals. This paper proposes a method which combines an overhearing and a segmented file delivery method based …

Yuki Koizumi

Yuki Koizumi

Osaka University

IEICE Technical Report; IEICE Tech. Rep.

Evaluation of Large-capacity Content Transmission Equipment to Bundle12 and 21GHz-band Satellite Transmission Channels and IP Network--Laboratory Experiment for Flexible …

(in English) We have developed large-capacity content transmission equipment that bundles 21-and 12-GHz-band satellite channels and an IP network for providing immersive content such as volumetric content that is large capacity due to expanding spatial expressions. This equipment achieves a maximum transmission capacity of about 1 Gbps by transmitting a content to multiple transmission channels and network where the capacity and delay are different. For immersive media, we are developing video and audio formats and scene descriptions to facilitate AR/VR schemes and a variety of other content representations appropriate for any given viewer environment. Immersive content enables users to experience various viewer environments by producing content for each object. Therefore, we developed a function that specifies the transmission channels or network in accordance with the objects of the content …