National Tsing Hua University Research Articles

Structure of monolayer FeSe on SrTiO3 deciphered

Picoscale structural insight into superconductivity of monolayer FeSe/SrTiO3 Authors: Rui Peng, Ke Zou, M. G. Han, Stephen D. Albright, Hawoong Hong, Claudia Lau, H. C. Xu, Yimei Zhu, F. J. Walker, and C. H. Ahn Sci. Adv. 6, eaay4517 (2020); DOI: 10.1126/sciadv.aay4517 Recommended with a commentary by Atsushi Fujimori, National Tsinghua University View Commentary (pdf)| […]

Salvage boron neutron capture therapy for pediatric patients with recurrent diffuse midline glioma.

Pediatric diffuse malignant glioma located in the brainstem was officially named "diffuse midline glioma" (DMG) by the World Health Organization in 2016. For this disease, radical surgery is not beneficial, and the only major treatment strategy is radiotherapy. However, the dose limitations to brainstem tissue mean that treatment by radiotherapy can only control and not eradicate the tumors, and there is no effective treatment for recurrence, resulting in short overall survival of 6-12months. This paper reports our experience with boron neutron capture therapy (BNCT), a new treatment process, and its efficacy in treating children with recurrent DMG. From September 2019 to July 2022, we treated 6 children affected by recurrent DMG. With the collaboration of Taipei Veteran General Hospital (TVGH) and National Tsing-Hua University (NTHU), each patient received two sessions of BNCT within 1month. Among the six patients, three showed partial response and the rest had stable disease after the treatment. The overall survival and recurrence-free survival duration after treatment were 6.39 and 4.35months, respectively. None of the patients developed severe side effects, and only one patient developed brain necrosis, which was most likely resulted from previous hypofractionated radiotherapy received. BNCT elicited sufficient tumor response with low normal tissue toxicity; it may benefit vulnerable pediatric patients with DMG.

‘“Thinking the Republic of China”: An International Symposium’

Abstract In the summer of 2021, scholars working in seven different countries gathered online to hold a series of seven different forums, each dedicated to discussing one aspect of National Tsinghua University professor Rur-bin Yang’s upcoming scholarly monograph Thinking the Republic of China (Forthcoming). The forum engaged with the following questions: what are the ideals of the Republic of China (roc)? How can one understand the meaning of the roc in relation to both the longue durée of imperial Chinese history and its once ideological adversary the People’s Republic of China? How can one historically evaluate the accomplishments of the roc in its post-1949 guise as a state-in-exile on the island of Taiwan? Has the combination of Confucian humanism and liberal constitutionalism imagined by early architects of the roc been realised in Taiwan, and do these ideals still have meaning for the larger Chinese world as a whole?

Two-Stream Spatial-Temporal Graph Convolutional Networks for Driver Drowsiness Detection.

Convolutional neural networks (CNNs) have achieved remarkable performance in driver drowsiness detection based on the extraction of deep features of drivers' faces. However, the performance of driver drowsiness detection methods decreases sharply when complications, such as illumination changes in the cab, occlusions and shadows on the driver's face, and variations in the driver's head pose, occur. In addition, current driver drowsiness detection methods are not capable of distinguishing between driver states, such as talking versus yawning or blinking versus closing eyes. Therefore, technical challenges remain in driver drowsiness detection. In this article, we propose a novel and robust two-stream spatial-temporal graph convolutional network (2s-STGCN) for driver drowsiness detection to solve the above-mentioned challenges. To take advantage of the spatial and temporal features of the input data, we use a facial landmark detection method to extract the driver's facial landmarks from real-time videos and then obtain the driver drowsiness detection result by 2s-STGCN. Unlike existing methods, our proposed method uses videos rather than consecutive video frames as processing units. This is the first effort to exploit these processing units in the field of driver drowsiness detection. Moreover, the two-stream framework not only models both the spatial and temporal features but also models both the first-order and second-order information simultaneously, thereby notably improving driver drowsiness detection. Extensive experiments have been performed on the yawn detection dataset (YawDD) and the National TsingHua University drowsy driver detection (NTHU-DDD) dataset. The experimental results validate the feasibility of the proposed method. This method achieves an average accuracy of 93.4% on the YawDD dataset and an average accuracy of 92.7% on the evaluation set of the NTHU-DDD dataset.

A Cloud-Based Model for Driver Drowsiness Detection and Prediction Based on Facial Expressions and Activities

This paper proposes an efficient approach to detecting and predicting drivers' drowsiness based on the cloud. This work focuses on the behavioral as well as facial expressions of the driver to detect drowsiness. This paper proposes an efficient approach to predicting drivers' drowsiness based on facial expressions and activities. Four different models with distinct features were experimented upon. Of these, two were VGG and the others were CNN and ResNet. VGG models were used to detect the movement of lips (yawning) and to detect facial behavior. A CNN model was used to capture the details of the eyes. ResNet detects the nodding of the driver. The proposed approach also exceeds the results set by the benchmark mode and provides high accuracy, an easy-to-use framework for embedded devices in real-time drowsiness detection. To train the proposed model, the authors have used the National Tsing Hua University (NTHU) Drivers Drowsiness data set. The overall accuracy of the proposed approach is 90.1%.

A Driver Drowsiness Detection Scheme Based on 3D Convolutional Neural Networks

It is an obvious fact that drivers’ drowsiness is more likely to cause traffic accidents. Recently, driver drowsiness detection has drawn considerable attention. In this paper, a novel drowsiness detection scheme is proposed, which can recognize drivers’ drowsiness actions through their facial expressions. First, a drowsiness action recognition model based on 3D-CNN is proposed, which can effectively distinguish drivers’ drowsiness actions and nondrowsiness actions. Second, a fusion algorithm of the two input streams is proposed, which can fuse gray image sequence and optical image sequence containing target motion information. Finally, the proposed model is evaluated on National Tsinghua University Driver Drowsiness Detection (NTHU-DDD) dataset. The experimental results show that the algorithm performs better than other algorithms, and its accuracy reaches 86.64%.

Preface

2021 The 7th International Conference on Digital Manufacturing and Automation (ICDMA 2021) was held from May 10-12, 2021 with workshop of 2021 6th International Conference on Mechanical and Electronics Engineering. Our attendees safty would be the highest priority, some participants are now facing travel restrictions and may have to cancel their schedule because of COVID-19. According to this situation, the conference is available to use all social media methods, such as online/video presentation, etc, except the regular presentation types (oral & poster).The conference is co-sponsored by National Tsing-Hua University, Taiwan and Royal Melbourne Institute of Technology, Australia, Science and Engineering Institute. It aims to bring together researchers, developers, and users in both industry and academia in the world for sharing state-of-art results, for exploring new areas of research and development, and to discuss emerging issues on digital manufacturing and automation. As is well known to us all, digital manufacturing and automation technology plays a more and more important role in industry advancing now, The conference is aimed at reporting up-to-the-minute innovations and developments, brought together leading researchers, engineers and scientists in the domain of interest from around the world. For conference online, three keynote speakers were invited to give the keynote speeches on May 11, 2021.This proceeding including 12 papers in two chapters of material design and optimization and industrial automation and manufacturing, the authors are from varies countries, hope those papers can help you to gain something and promote the development of Digital Manufacturing and Automation and Mechanical and Electronics Engineering. The conference received 23 submissions, and every paper was reviewed by in two roundsFor the success helding of ICDMA 2021, we would like to express our appreciation to all the reviewer's review works and their support, all the conference website editor and committee members' hard works. It very nice to have you worked for conference ICDMA 2021. The conference would welcome to your join for next conference series.At last, wish you could get something from our conference and conference proceedings!!Conference Committees, Peer Review Statement, are available in this pdf

Methodology for analyzing accidents with radioactive material release with code EPZDose

Abstract A methodology for analyzing accidents with radioactive material release with EPZDose code was established. This code assesses doses and it is designed and developed by NTHU (National Tsing Hua University). To confirm the capacity of EPZDose, three postulated accident scenarios Taiwanese NPPs Chinshan (BWR/4) and Maanshan (PWR) are analyzed. All these scenarios are SBO (station blackout) transients because it is assumed that they result in a release of radioactive material. In this study, the source term data for EPZDose are taken from MELCOR or RASCAL calculations. In addition, calculated results of RASCAL code are compared with the results of EPZDose for these scenarios. The comparison show that the EPZDose predictions are consistent with the data of RASCAL. This indicates that the EPZDose has a respectable accuracy in the analysis of radioactive material release accidents.

The Last Byte: Baseball and Testing

If you are curious about the title, you may not be alone. When I was invited to give a keynote speech at the 2019 International Test Conference in Asia (ITC-Asia), I was helping my colleagues at National Tsing Hua University (NTHU) establish the Center for Sport Science and Technology. So I thought, why not combine the two things that have shaped my entire life? The United States, which has a vision of “all children have the opportunity to be active through sports” and a policy of “sport for all, play for life,” is where baseball was born. And, of course, we all know it is where transistors and integrated circuits were invented. Therefore, there may be more people than we know who live in the two worlds simultaneously. I do. I played in the 1971 Little League World Series, and when I mentioned this once, Rob Aitken immediately disclosed that he also played Little League Baseball (LLB). Bill Lowd in the Steering Committee of ITC is also an administrator of LLB. Today, it is becoming clear that the two worlds have more to share than before.

Academic Entrepreneurship for Social Innovation in Taiwan: The Cases of the OurCityLove Platform and the Forest App

This study attempts to extend the definition of social innovation within the context of academic entrepreneurship. We consider how academic entrepreneurs can undertake commercial activities, and which ones, and how these activities contribute to the contexts of social innovation. We explore two cases that are derived from two premier universities in Taiwan in terms of research and academic entrepreneurship: OurCityLove from National Chiao Tung University (NCTU) and the Forest app from National Tsing Hua University (NTHU). The two cases show how social enterprises achieved the financial ability to expand their businesses and create the desired social values. While the first case, derived from NCTU, focused on providing useful information on social spaces, and services for the elderly, parents traveling with babies and those with disabilities (and also creating job opportunities for the latter), the other from NTHU created an app to influence those addicted to playing with their smartphones. The cases illustrate how the two universities capitalised on their technological competencies and academic programmes to support graduates and researchers in venturing into social entrepreneurship.

Non-intrusive Driver Drowsiness Detection based on Face and Eye Tracking

The rate of annual road accidents attributed to drowsy driving are significantly high. Due to this, researchers have proposed several methods aimed at detecting drivers' drowsiness. These methods include subjective, physiological, behavioral, vehicle-based, and hybrid methods. However, recent reports on road safety are still indicating drowsy driving as a major cause of road accidents. This is plausible because the current driver drowsiness detection (DDD) solutions are either intrusive or expensive, thus hindering their ubiquitous nature. This research serves to bridge this gap by providing a test-bed for achieving a non-intrusive and low-cost DDD solution. A behavioral DDD solution is proposed based on tracking the face and eye state of the driver. The aim is to make this research an inception to DDD pervasiveness. To achieve this, National Tsing Hua University (NTHU) Computer Vision Lab's driver drowsiness detection video dataset was utilized. Several video and image processing operations were performed on the videos so as to detect the drivers' eye state. From the eye states, three important drowsiness features were extracted: percentage of eyelid closure (PERCLOS), blink frequency (BF), and Maximum Closure Duration (MCD) of the eyes. These features were then fed as inputs into several machine learning models for drowsiness classification. Models from the K-nearest Neighbors (KNN), Support Vector Machine (SVM), Logistic Regression, and Artificial Neural Networks (ANN) machine learning algorithms were experimented. These models were evaluated by calculating their accuracy, sensitivity, specificity, miss rate, and false alarm rate values. Although these five metrics were evaluated, the focus was more on getting optimal accuracies and miss rates. The result shows that the best models were a KNN model when k = 31 and an ANN model that used an Adadelta optimizer with 3 hidden layer network of 3, 27, and 9 neurons respective. The KNN model obtained an accuracy of 72.25% with a miss rate of 16.67%, while the ANN model obtained 71.61% and 14.44% accuracy and miss rate respectively.

A Reliability Look at Energy Development

Professor Way Kuo, President of City University of Hong Kong, previously served on the senior management team at Oak Ridge National Laboratory and as Dean of Engineering at the University of Tennessee, and Head of Industrial Engineering at Texas A&M University. A pioneer in reliability, he is renowned for his work in design for reliability of electronics and energy systems. He is a member of the US National Academy of Engineering and Academia Sinica in Taiwan and a Foreign Member of the Chinese Academy of Engineering and the Russian Academy of Engineering. Dr. Chin Pan is Tsing Hua distinguished professor of the Institute of Nuclear Engineering and Science and the Department of Engineering and System Science of the National Tsing Hua University (NTHU), Taiwan. Dr. Pan previously served as Dean of the College of Nuclear Science of NTHU and Convener of the Energy Program of the Ministry of Science and Technology of Taiwan. Dr. Pan's research activities have been in the areas of two-phase flow, boiling heat transfer, nuclear thermal hydraulics, and energy engineering.

Corrigendum: Triggering mitophagy with far-red fluorescent photosensitizers.

Scientific Reports 5: Article number: 10376; published online: 29 May 2015; updated: 10 June 2016 In the original version of this Article, an additional affiliation for Cheng-Wei Hsieh was omitted. The correct affiliation is listed below: Department of Chemistry, National Tsing-Hua University, Hsinchu 30013, Taiwan.

Special issue dedicated to the memory of Paul Waltman

This volume is dedicated to the memory of Paul Waltman. Many of the authors of articles contained here were participants at the NCTS International Conference on Nonlinear Dynamics with Applications to Biology held May 28-30, 2014 at National Tsing-Hua University, Hsinchu, Taiwan. The purpose of the conference was to survey new developments in nonlinear dynamics and its applications to biology and to honor the memory of Professor Paul Waltman for his influence on the development of Mathematical Biology and Dynamical Systems. Attendees at the conference included Paul's sons Fred and Dennis, many of Paul's former doctoral and post-doctoral students, many others who, although not students of Paul, nevertheless were recipients of Paul's valuable advice and council, and many colleagues from all over the world who were influenced by Paul's mathematics and by his personality. We thank the NCTS for its financial support of the conference and Dr. J.S.W. Wong for supporting the conference banquet. For more information please click the “Full Text” above.

Crack-free Micromachining of Glass Ceramic Using Visible LIBWE

Research Center for Applied Sciences, Academia Sinica Taiwan, Taipei City, 11529, Taiwan ROT Institute of Biophotonics, National Yang-Ming University, Taipei City 11221, Taiwan Molecular Imaging Research Center, National Yang-Ming University, Taipei City 11221, Taiwan Department of Mechanical and Mechatronic Engineering, National Taiwan Ocean University, Keelung 20224, Taiwan Department of Engineering and System Science, National Tsing-Hua University, Taiwan Nano Science and Technology Program, Taiwan International Graduate Program, Academia Sinica Taiwan Department of Chemistry, National Taiwan University, Taiwan

A compact biological cell irradiation system with a Van de Graaff accelerator

This report describes the development of a compact microcell irradiation system in the 3MV KN Van de Graaff accelerator in the Accelerator Laboratory of National Tsing Hua University (NTHU). The main feature of this system is backscattering, a technique whereby a 100nm gold scattering foil is placed in the center of a scattering chamber, instead of a 90° bending magnet. The incident particles produced by the accelerator bombard the scattering foil and scatter isotropically. A set of micro pinholes was installed above the scattering foil and directed the 90° scattering particles to irradiate the cell target. This innovation simplifies the system design and massively reduces the cost and space requirements for system construction.

BNCT treatment planning of recurrent head-and-neck cancer using THORplan

A cooperation program on Boron Neutron Capture Therapy (BNCT) between National Tsing Hua University (NTHU) and Taipei Veterans General Hospital (TPEVGH) was established in 2008. Clinical trial of recurrent head-and-neck cancer is the goal of the program. In this study, treatment plannings of two head-and-neck cancer cases are performed using treatment planning system THORplan developed at NTHU of Taiwan. The patients are assumed to be irradiated under current THOR epithermal neutron beam. The prescription dose is 20 Gy-Eq for at least 80% of tumor volume. The irradiation time to reach the target tumor dose can be kept within 1h. The skin dose is within the limiting dose of 11 Gy-Eq. The spinal cord dose is well within the limiting dose of 10 Gy-Eq. The use of an extension collimator for easier patient positioning is helpful in reducing the dose of eye lens to within the dose limit of 5 Gy-Eq. The irradiation time, however, will increase slightly due to the increase of source-to-tumor distance. The CPU time for treatment planning calculation is ~10 h. With the use of user friendly treatment planning system THORplan, dose planning for BNCT at THOR can be easily performed.

The Center for Nanotechnology, Materials Science, and Microsystems

The center for Nanotechnology, Materials Science, and Microsystems (CNMM) at the National Tsing Hua University (NTHU) was established in 2006, founded by the National Science Council (NSC). It consists of three parts, including core-facility clean rooms, the Nanomeasurement Laboratory, and a metallurgical factory. They work synergjstically to serve faculties, graduate students, and scientists within and outside the campus for their research needs.

Nanoscience and Nanotechnology [The Editor's Desk

This is going to be an exciting year for R&D activities in Asia as speculative “hot money” flows into China, some of which will be used to advance China's manufacturing capability to a higher level. This issue starts with an introductory article from the Center for Nanotechnology, Materials Science and Microsystems (CNMM) of National Tsing Hua University (NTHU).

Highly Alpha‐Selective Sialyl Phosphate Donors for Efficient Preparation of Natural Sialosides

Hubble, bubble, toil, and trouble: The use of a new sialyl phosphate donor allows the stereoselective, one-pot multicomponent synthesis of α-sialooligosaccharides (see scheme).