检索结果-内蒙古大学图书馆

Environmental microorganism classification using optimized deep learning model

ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH 2021年第24期28卷 31920-31932页

作者： Liang, Chih-Ming Lai, Chun-Chi Wang, Szu-Hong Lin, Yu-Hao Feng Chia Univ Dept Environm Engn & Sci 100 Wenhwa Rd Taichung 40724 Taiwan Natl Yunlin Univ Sci & Technol Bachelor Program Interdisciplinary Studies 123 Univ RdSect 3 Touliu 64002 Yunlin Taiwan China Med Univ Coll Humanities & Sci Master Program Digital Hlth Innovat 100Sect 1Jingmao Rd Taichung 406040 Taiwan China Med Univ Ctr Gen Educ 100Sect 1Jingmao Rd Taichung 406040 Taiwan

Rapid environmental microorganism (EM) classification under microscopic images would help considerably identify water quality. Because of the development of artificial intelligence, a deep convolutional neural network (CNN) has become a major solution for image classification. Three popular CNNs, referred to as ResNet50, Vgg16, and Inception-v3, were transferred to identify the EM images present on the Environmental Microorganism Dataset (EMDS), and EMAD was the small dataset, which only has 294 EM images with 21 EM classes. Besides data augmentation, optimizing the fully connected layer of CNN, i.e., both optimally fine-tuned neuron number and dropout rate, was adopted to enhance the performance produced by CNN. The discussions on the causes of the accuracy improved by optimization are also provided. The results showed that the Inception-v3 model obtained 84.9% of the accuracy and performed better than the other two famous CNNs. Also, the implement of data augmentation enhanced the performance of Inception-v3 on EMDS. To add to that, the optimized Inception-v3 model archived 90.5% of the accuracy, and this result demonstrated the improvement effect obtained by using genetic algorithm (GA) to optimize the fully connected layer of the Inception-v3. Therefore, the optimize Inception-v3 with data augmentation process obtained the accuracy of 92.9% and improved almost 21% higher than that obtained from the famous Vgg16. In addition, the optimized Inception-v3 would need less neurons, when compared with that of the optimized Vgg16 possibly. This optimized Inception-v3 could provide a solution to the EM classification in microscope with a digital camera system.

关键词： Deep learning Genetic algorithm (GA) Convolutional neural network (CNN) Transfer learning Environmental microorganism Artificial intelligence (AI)

来源：评论

学校读者我要写书评

暂无评论

Magnetic susceptibility, optical, and adhesive properties of Co40Fe40V10B10 films

引用

SURFACE ENGINEERING 2020年第4期36卷 405-410页

作者： Liu, Wen-Jen Ou, Sin-Liang Chang, Yung-Huang Chen, Yuan-Tsung Wang, Yu-Tang Liang, You-Cheng Tseng, Jiun-Yi Ch, Po-Wei I Shou Univ Dept Mat Sci & Engn Kaohsiung Taiwan Da Yeh Univ Bachelor Program Design & Mat Med Equipment & Dev Changhua Taiwan Natl Yunlin Univ Sci & Technol Bachelor Program Interdisciplinary Studies Touliu Taiwan Natl Yunlin Univ Sci & Technol Grad Sch Mat Sci 123 Univ RdSect 3 Touliu 64002 Yunlin Taiwan Acad Sinica Inst Phys Taipei Taiwan

The low-frequency alternating current magnetic susceptibility (chi(ac)) property decreased when the frequency increased. The highest chi(ac) meant the spin sensitivity was maximized at an f(res). Of all CoFeVB thicknesses, the 20-mm thickness had the highest chi(ac) value (9.85 x 10(-3)) at an f(res) of 1000 Hz. The transmittance percentage decreased from 43% to 10% as the thicknesses changed from 10 to 40 nm. The absorbance of the CoFeVB films decreased. Finally, the surface energy revealed the lowest value at 20 nm. A high surface energy indicates high adhesive characteristics. The maximal surface energy of the 40-nm thickness was 38 mJ mm(-2), indicating excellent adhesion. The optimal magnetic performance of 20-nm film, because it exhibits the highest chi(ac) value.

关键词： Cobalt iron vanadium boron alloy low-frequency alternate-current magnetic susceptibility resonance frequency transmittance percentage adhesion

来源：评论

学校读者我要写书评

暂无评论

MULTI-LABEL CLASSIFICATION OF ICD CODING USING DEEP LEARNING

MULTI-LABEL CLASSIFICATION OF ICD CODING USING DEEP LEARNING

引用

International Symposium on Community-Centric Systems (CcS)

作者： Hsu, Chung-Chian Chang, Pei-Chi Chang, Arthur Natl Yunlin Univ Sci & Technol Dept Informat Management Touliu Yunlin Taiwan Natl Yunlin Univ Sci & Technol Bachelor Program Interdisciplinary Studies Touliu Yunlin Taiwan

ISBN: (纸本)9781728187419

This study uses deep learning approach to tackle the multi-label classification problem in ICD coding. The discharge summaries on MIMIC-III dataset are adopted to explore the training methods of text preprocessing, label preprocessing, and model training. Specifically, the methods of label-to-chapter and common label classification are experimented, hoping to find the best recommendations of ICD codes for each diagnosis by the physicians. The result shows CNN has the best performance 76% by micro F1-measure in label-to-chapter. In addition, it also shows CNN outperforms other methods within top-50 and top-100 common labels.

关键词： ICD coding multi-label classification natural language processing deep learning

来源：评论

学校读者我要写书评

暂无评论

The effects of an augmented-reality ubiquitous writing application: a comparative pilot project for enhancing EFL writing instruction

引用

COMPUTER ASSISTED LANGUAGE LEARNING 2022年第5-6期35卷 989-1030页

作者： Lin, Vivien Liu, Gi-Zen Chen, Nian-Shing Natl Yunlin Univ Sci & Technol Bachelor Program Interdisciplinary Studies Coll Future Touliu Yunlin Taiwan Natl Yunlin Univ Sci & Technol Dept Appl Foreign Languages Touliu Yunlin Taiwan Natl Cheng Kung Univ Dept Foreign Languages & Literature Tainan Taiwan

The use of technology such as online software has been examined in English as a foreign language (EFL) writing contexts. However, few studies have incorporated context-aware ubiquitous technology into EFL writing instruction. To develop multimodal and digital literacy of target EFL undergraduates in this pilot project, the researchers designed and implemented an augmented-reality context-aware ubiquitous writing (ARCAUW) application, which aimed to increase long-term memory, motivation, and self-regulated cognition in participants' writing development. The pilot project compared the writing outcomes and learner perceptions of the proposed ARCAUW writing mode against the baseline mobile-assisted, classroom-based writing mode. The two writing modes differed in the way that metacognitive scaffolding tasks were carried out. Pre- and post-test results showed that (a) although both modes led to significant improvement in writing the process analysis essay, ARCAUW was conducive to the development of task schema in long-term memory, motivation, and self-regulation in writing, and (b) additional cognitive processing during AR-based learning led to mixed results in writing performance. It was concluded that ARCAUW should be incorporated into mobile-assisted writing courses to reach optimal instructional outcomes. Finally, a five-step procedure was offered to help EFL practitioners design and implement ubiquitous writing in their own teaching contexts.

关键词： English as a foreign language writing augmented-reality context-aware ubiquitous learning genre-based writing instruction metacognitive scaffolding self-regulated writing

来源：评论

学校读者我要写书评

暂无评论

Development of a Smart Multifunctional Visitor-Identification System 8

Development of a Smart Multifunctional Visitor-Identificatio...

引用

8th IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW) - Driving World to 5G and Beyond with Consumer Technology

作者： Chen, Chih-Chen Lue, Ren-Guang Hsieh, Tsung-Hsien Hwang, Chi-Pan Chen, Yu-Luen Wang, Hsuan-Fu Hwa Hsia Univ Technol New Taipei Taiwan Natl Changhua Univ Educ Changhua Changhua County Taiwan Natl Taipei Univ Educ Taipei Taiwan Natl Yunlin Univ Sci & Technol R&D Ctr Intelligent Elect Prod Bachelor Program Interdisciplinary Studies Touliu Yunlin Taiwan

ISBN: (纸本)9781665433280

Our article establishes a mechanism to remind residents whether there are guests visiting through related electronic devices for range finding, identification, etc. A vibrating bracelet (or ring) and a flashing light bar are worn on the resident's wrist (or neck), that is, to remind the resident by means of vibration and light bar flashing. Other related hardware aspects include LED dot matrix modules, voice broadcast, and infrared door locks;LED dot matrix modules are used to display the current date, time, and visitor information, and the voice broadcast will play the speakers of the visiting guests. The LED dot matrix module can provide information about visiting guests for the hearing impaired or residents with reduced mobility, and the voice broadcast can provide information about the visiting guests for the visually impaired or residents with reduced mobility. If the resident completes the identification of the visiting guest and wants to open the door lock, he can use infrared light wave to open the door lock. This is a convenient and safe way for residents with limited mobility or the elderly.

关键词： Wrist Vibrations Conferences Senior citizens Auditory system Light emitting diodes Hardware

来源：评论

学校读者我要写书评

暂无评论

Identification of Fruit Tree Pests With Deep Learning on Embedded Drone to Achieve Accurate Pesticide Spraying

引用

IEEE ACCESS 2021年 9卷 21986-21997页

作者： Chen, Ching-Ju Huang, Ya-Yu Li, Yuan-Shuo Chen, Ying-Cheng Chang, Chuan-Yu Huang, Yueh-Min Natl Yunlin Univ Sci & Technol Dept Bachelor Program Interdisciplinary Studies Touliu 64002 Yunlin Taiwan Natl Cheng Kung Univ Dept Engn Sci Tainan 70101 Taiwan Tainan Dist Agr Res & Extens Stn Div Crop Environm Tainan 712009 Taiwan Natl Yunlin Univ Sci & Technol Dept Comp Sci & Informat Engn Touliu 64002 Yunlin Taiwan

Tessaratoma papillosa (Drury) first invaded Taiwan in 2009. Every year, T. papillosa causes severe damage to the longan crops. Novel applications for edge intelligence are applied in this study to establish an intelligent pest recognition system to manage this pest problem. We used a detecting drone to photograph the pest and employed a Tiny-YOLOv3 neural network model built on an embedded system NVIDIA Jetson TX2 to recognize T. papillosa in the orchard to determine the position of the pests in real-time. The pests' positions are then used to plan the optimal pesticide spraying route for the agricultural drone. Apart from planning the optimized spraying of pesticide for the spraying drone, the TX2 embedded platform also transmits the position and generation of pests to the cloud to record and analyze the growth of longan with a computer or mobile device. This study enables farmers to understand the pest distribution and take appropriate precautions in real-time. The agricultural drone sprays pesticides only where needed, which reduces pesticide use, decreases damage to the environment, and increases crop yield.

关键词： Drones Agriculture Image edge detection Real-time systems Deep learning Computational modeling Object detection Edge intelligence unmanned aerial vehicles (UAV) real-time embedded systems slope land orchard object detection agricultural pests damage precision agriculture intelligent pest recognition

来源：评论

学校读者我要写书评

暂无评论

Forming a TED talks sphere for convenient search

引用

ELECTRONIC LIBRARY 2020年第2期38卷 403-420页

作者： Shih, Chia-Hui Li, Han-Lin Hu, Chih-Chien Lin, Bertrand M. . Natl Chiao Tung Univ Inst Informat Management Hsinchu Taiwan Natl Yunlin Univ Sci & Technol Coll Future Bachelor Program Interdisciplinary Studies Touliu 64002 Yunlin Taiwan

Purpose TED (Technology, Entertainment, Design, ) Talks has been one of the most popular video systems. However, the current TED Talks system expressed its inquired videos as in a two-dimensional (2D) table, which is inconvenient for searching the relationships among videos and tags. This study converts the TED Talks table into a sphere by using optimization techniques to help users search for preferred videos. Design/methodology/approach There are five phases in this study as follows. Phase 1: Reorganize data of 36 tags and 108 videos;Phase 2: Allocate tags on the TED sphere;Phase 3: Allocate videos on the TED sphere;Phase 4: Develop an online interactive TED retrieval system;and Phase 5: Perform survey and evaluation. Findings One survey demonstrated that the TED Talks sphere is more convenient for searching videos, as it is more user-friendly because of its graphical user interface, more convenient to use, more useful for retrieving information and can facilitate a more responsive search for users' preferred videos. Research limitations/implications The numbers of tags and videos able to be displayed on a sphere is limited by the capacity of an optimization software and hardware. Practical implications The proposed sphere system can be used by a large number of users of TED Talks groups. Social implications This sphere systems can also be applied to other fields which use 2D forms to display the relationships among objects. Originality/value This study uses an optimization method to convert a 2D form into a 3D sphere to highlight the relationships among numerous objects.

关键词： Optimization TED Talks Visualization 3D sphere Retrieval systems Databases

来源：评论

学校读者我要写书评

暂无评论

Effects of Interfacial Oxidative Layer Removal on Charge Carrier Recombination Dynamics in InP/ZnSexS1-x Core/Shell Quantum Dots

引用

JOURNAL OF PHYSICAL CHEMISTRY LETTERS 2021年第30期12卷 7194-7200页

作者： Pu, Ying-Chih Fan, Hsiao-Chuan Chang, Jui-Cheng Chen, Yu-Hung Tseng, Shih-Wen Natl Univ Tainan Dept Mat Sci Tainan 70005 Taiwan Natl Yunlin Univ Sci & Technol Dept Chem & Mat Engn Touliu 64002 Yunlin Taiwan Natl Yunlin Univ Sci & Technol Bachelor Program Interdisciplinary Studies Touliu 64002 Yunlin Taiwan Natl Cheng Kung Univ Coll Med Dept Med Tainan 70101 Taiwan Natl Cheng Kung Univ Core Facil Ctr Tainan 70101 Taiwan

Red-light-emitting InP/ZnSexS1-x core/shell quantum dots (QDs) were prepared by one-pot synthesis with optimal hydrogen fluoride (HF) treatment. Most of the surficial oxidative species could be removed, and the dangling bonds would be passivated by Zn ions for the InP cores during HF treatment, which would be beneficial to the subsequent ZnSexS1-x shell coating. Three-dimensional time-resolved photoluminescence spectra of the QD samples were analyzed by singular value decomposition global fitting to determine the radiative and nonradiative lifetimes of charge carriers. A proposed model illustrated that the charge carriers in the InP/ZnSexS1-x QDs with interfacial oxidative layer removal would evidently recombine through radiative pathways, mainly from the conduction band to the valence band (lifetime, 33 ns) and partially from the trap states (lifetime, 150 ns). This work offers the important physical insight into the charge carrier dynamics of low-toxicity QDs which have the desired optical properties for optoelectronic applications.

关键词： Quantum dots Layers Carrier dynamics Electrical conductivity Recombination

来源：评论

学校读者我要写书评

暂无评论

Unnecessary risks created by uncontrolled rocket reentries

arXiv

引用

arXiv 2022年

作者： Byers, Michael Wright, Ewan Boley, Aaron Byers, Cameron Department of Political Science University of British Columbia VancouverBC Canada Interdisciplinary Studies Graduate Program University of British Columbia VancouverBC Canada Department of Physics and Astronomy University of British Columbia VancouverBC Canada Bachelor of Engineering Program University of Victoria VictoriaBC Canada

In 2020, over 60% of launches to low Earth orbit resulted in one or more rocket bodies being abandoned in orbit and eventually returning to Earth in an uncontrolled manner. When they do so, between 20 and 40% of their mass survives the heat of atmospheric reentry. Many of the surviving pieces are heavy enough to pose serious risks to people, on land, at sea, and in airplanes. There is no international consensus on the acceptable level of risk from reentering space objects. This is sometimes a point of contention, such as when a 20 tonne Long March 5B core stage made an uncontrolled reentry in May 2021. Some regulators, including the US, France, and ESA, have implemented a 1 in 10, 000 acceptable casualty risk (i.e., statistical threat to human life) threshold from reentering space objects. We argue that this threshold ignores the cumulative effect of the rapidly increasing number of rocket launches. It also fails to address low risk, high consequence outcomes, such as a rocket stage crashing into a high-density city or a large passenger aircraft. In the latter case, even a small piece could cause hundreds of casualties. Compounding this, the threshold is frequently ignored or waived when the costs of adherence are deemed excessive. We analyse the rocket bodies that reentered the atmosphere from 1992-2021 and model the associated cumulative casualty expectation. We then extrapolate this trend into the near future (2022-2032), modelling the potential risk to the global population from uncontrolled rocket body reentries. We also analyse the population of rocket bodies that are currently in orbit and expected to deorbit soon, and find that the risk distribution is significantly weighted to latitudes close to the equator. This represents a disproportionate burden of casualty risk imposed on the countries of the Global South by major spacefaring countries. Modern rockets have reignitable engines, allowing controlled reentries into remote areas of the oceans. This, combine

关键词： Rockets

来源：评论

学校读者我要写书评

暂无评论

A Real-Time Monocular Vision-Based Obstacle Detection 6

A Real-Time Monocular Vision-Based Obstacle Detection

引用

6th International Conference on Control, Automation and Robotics (ICCAR)

作者： Wang, Szu-Hong Li, Xiang-Xuan Natl Yunlin Univ Sci & Technol Bachelor Program Interdisciplinary Studies Touliu 640 Yunlin Taiwan Natl Yunlin Univ Sci & Technol Dept Elect Engn Touliu 640 Yunlin Taiwan

ISBN: (纸本)9781728161396

In recent years, many robotic obstacle avoidance methods use high-priced laser detection technology, or a dual-lens camera as a sensor for robot control. In order to reduce costs, this paper presents a real-time obstacle detection system by using a single wide-angle camera. In the robot application, the system can detect various static or dynamic objects in front of the robot body. The proposed system uses a local object based background subtraction method to get the outline of the static/moving object. The adaptive integral image based obstacle detection method without manual adjustment of setting parameters. The obstacle detection system had been implemented by ARM system, the obstacle can be detected on the real-time camera. The frame rate is close to 32 fps with Raspberry pi 3 platform.

关键词： obstacle detection robot monocular vision

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：