Chung Cheng University Institutional Repository
Not a member yet
889 research outputs found
Sort by
論表演人之權利保護 -以著作人格權為中心;
No full text[[abstract]]為了因應科技發展的腳步，世界智慧財產權組織於2012年通過視聽表演北京條約，使得表演人之權利保護更加完整。本文從表演人保護之緣起出發，介紹表演人之定義與範圍、表演人權利保護之法律性質，接著從各國際公約、各國立法例以及我國現行著作權法與我國著作權法修正草案中，完整介紹國際間及我國對表演人權利保護包括著作人格權與著作財產權之具體實踐，並將焦點關注在表演人之著作人格權，透過比較法與判決之研究，檢討我國表演人之著作人格權的保護。最後進一步從三個面向出發：第一、從契約自治的角度著手約束契約雙方；第二、討論以合理使用限制著作人格權之可能性；第三、從時間的限制著手，討論著作人格權保護期間，試圖找出限制表演人之著作人格權之方法。
In response to the pace of technological development, the World Intellectual Property Organization in 2012 signed the Beijing Treaty on Audiovisual Performances to provide a complete protection of the performers’ right. This research starts from the origin of the protection of the performers. Furthermore, the author introduces the protection of the performers’ right in the international convention, the legislation of various nationals, and our present legislation and future legislation amendment. In this research, the author focus on moral rights of the performers. By using comparative study method and cases study, the author reviews the protection of moral rights of the performers. Finally, from the point of view of fair use, contractual freedom and copyright protection term, the author tries to find out the way to limit the moral rights of the performers
基礎核心訓練和彈力帶運動訓練對社區高齡者功能性體適能之影響;Effects of Foundation Training and elastic band exercise training on Functional Fitness Performance in Community-Dwelling Older Adults
No full text[[abstract]]背景：人口老化已經成為全球發展中國家的趨勢，但老化卻帶來高齡者生理機能退化、慢性疾病罹患率增高以及跌倒風險的增高等問題；進而造成高齡者日常活功能能力的喪失。長壽卻不健康的高齡身體狀況，不論是在醫療或是社會成本的支出，儼然成為各國政府沉重的負擔。因此延緩老化所帶來的各種社會問題即成為所有已發展國家的重要議題。肌力訓練對於老年人極為重要，不僅能防止因為老化所引起的肌肉流失、低骨質密度和低心血管循環能力，也能加強高齡者的活動能力、平衡能力和步態。目的：本研究目的在於以預防醫學、運動科學與健康促進之結合，發展有效增進社區高齡者功能性體適能的基礎核心訓練運動和彈力帶運動課程。並且在未來提供給社區和機構之高齡者做為參考課程，已達健康促進之貢獻。方法：研究對象為65歲以上社區內健康的高齡者，介入8週基礎核心與彈力帶訓練課程，觀察社區高齡者在功能性體適能的表現。結果：一、基礎核心訓練課程介入對於社區高齡者在上、下肢肌肉力量的前後測有顯著進步；二、彈力帶訓練課程介入對於社區高齡者在下肢柔軟度與握力表現的前後測有顯著進步。結論：基礎核心運動和彈力帶運動訓練藉由全面性阻力運動設計介入能有效改善社區高齡者功能性體適能中的上肢肌力、下肢肌力與握力。期許未來將此安全且有效的全面性運動處方策略推廣至更多社區，讓更多久高齡族群生活型態變得更活躍，進而能獲得身心健康的提升。本研究之成果可作?未來老年人預防醫學的參考。
Background: The aging of the population has become the trend of developing countries in the world, but the aging has brought the elderly physiological function degradation, increased incidence of chronic disease and increased risk of falling; resulting in the loss of daily living function of the elderly. The elderly physical condition is longevity but not healthy, and the costs of medical or social has become a heavy burden on governments. So that the various social problems caused by aging are becoming important issues for all developed countries. Muscle strength is very important for elderly adults. Muscle strength is not only could decrease muscle mass lose, low density of bone mass and poor circulation of vein due to aging, but also improve ability of activity, balance and gait for elderly adults. Purposes: The purpose of this study that combine preventive medicine, exercise science and health promotion, and development of an effective foundation training and elastic band training program to promote the functional fitness of elderly people in the community. In the future, the course will become a reference to elderly in the community and institutions. There is the contribution of health promotion. Methods: The subjects were healthy middle-aged people aged over 65 years, 8 weeks of foundation training and elastic band training, observe the performance on the functional fitness in the elderly adults in the community. Results: 1. The intervention of foundation training course for the elderly in the community in the upper and lower muscle strength before and after the test has made significant progress. 2. The intervention of elastic band training course for the elderly in the community in the lower limb flexibility and grip strength pre and post-test has made significant progress.Conclusions: Through effective foundation training and elastic band training, integrated group resistanse exercise program training could enhance upper and lower body strength, and grip strength, in turn. These results suggest that foundation training and elastic band training are needed to promote increased in physical activity participation for more elderly communities for a bigger population of rural older adults. This study provides evidence for the future study of preventive medicine in rural community elderly
探討知識分享與開源軟體開發效能：以社會交換理論與社會資本理論觀點;Knowledge Sharing and OpenSource Software Development Research：from the Perspective of Social-Exchange Theory and Social-Capital Theory
No full text[[abstract]]?文摘要 在現今知識爆炸的現代社會中，資源共享已經是必然的趨勢，而在本研究中所提到的開源一詞，並不僅僅局限於軟體開發方面，開源更是一種生活態度，也代表了對於現今大眾對於公共事務冷漠的反思。 本研究採用網?問卷調查及蒐集樣本，並?結於開源軟體開發相關?壇、社群網站、 PTT Soft_job、Java等軟體開發版面。資?分析方法採結構化方程模式，並以 SmartPLS3.0 作為主要統計分析工具，驗證研究模式中個變?的因果關係。研究結果顯示，其中社會交換理論中的變數–互惠對於正向影響開源軟體開發社群間的知識分享行為較不明顯外，其餘之信任、權力以及社會資本理論中的社會互動連結、認同及共同願景皆會正向影響知識分享行為，而知識分享行為對於開源軟體開發效能上的影響也是正向顯著的。 鑑於知識分享已經是現今科技資訊應用上的主要議題，因此本研究試圖加入以社會資本、社會交換理論等心理學與社會現在觀察理論的關鍵因素，以探開源軟體開發過程與結果的影響基礎，透過以知識分享之中介變數，並找出其他可能影響相關人員，如開發人員、相關開發社群討論區使用者等知識分享的變數，試圖歸納並改進使用者的知識分享行為。
In today's knowledge explosion of modern society, the behavior of resource sharing is an inevitable trend, and in this research referred to the open source term, not just limited to software development, open source is a life attitude, but also represents the Today 's Public Reflection on Public Affairs. This research is based on the data of questionnaire collecting from internet. And also public the questionnaire link on shopping forums and network social media, The method of data analysis is based on the module equation. And use "SmartPLS3.0" as the major analysis tool to verify the relation of variables, The result of research reveal the variable of SET “Reciprocity” will not impact on the behavior of knowledge sharing directly, But the other variables of SET and SCT will do. And Also the KS impact the efficient of OSSD.In view of the fact that knowledge sharing is the main topic in the application of information technology today, this research attempts to combine the social capital theory(SCT)、social exchange theory(SET) and in psychology and social observation theory of the key factors in order to explore the source software development process and results of the foundation, This paper attempts to summarize and improve the knowledge-sharing behavior of the users through the mediating variables of KS and identify other variables that may influence the sharing of knowledge among relevant personnel, such as developers, discussion forum users and other related development communities
以多重觀點探討接受血糖行動健康系統之關鍵因素;Exploring the key factors in the acceptance of the mobile blood sugar monitor system from multiple perspectives.
No full text[[abstract]]隨著科技進步，健康管理結合行動科技，行動健康系統除須考量行動化特性外，需考量不同類型被照護者需求。在心理因子與代謝症候群關聯性研究中，控制傳統危險因子後發現心理因子與代謝症候群具正向關聯性，本研究認為血糖檢測使用者對血糖行動健康系統，其接受因素將不同於一般健康者。本研究整?過去文獻，從新科技、心理及個人層面觀點探討血糖檢測使用者接受血糖行動健康系統之關鍵因素。本研究共收集462 份有效問卷，透過SPSS 20.0 進行敘述性統計、信效度分析，使用Smart PLS 3.2 進行研究模型檢定及驗證模型解釋力。分析結果發現，從使用者接受新科技觀點，?績效期望?、?努力期望?、?社會影響?、?促成條件?、?習慣?會影響血糖檢測使用者接受血糖行動健康系統。而從使用者心理層面觀點，?相對優勢?會影響血糖檢測使用者接受血糖行動健康系統；?參考價格?、?參考處境?則不會有影響。從使用者的個人層面觀點探討，?自我效能?、?信任?、?風險偏好?、?健康意識?因素會影響血糖檢測使用者接受血糖行動健康系統。從使用者使用行為探討，則?促成條件?、?習慣?、「行為意圖」會影響血糖檢測使用者接受血糖行動健康系統。本研究根據資料分析結果，驗證解釋影響血糖檢測使用者接受血糖行動健康系統之關鍵因素。其結果期望能對相關學術界、產業界有所貢獻。
Healthcare management now is combined with mobile technologies. In the mobile health-monitoring system, along with the features of mobility, it is essential to consider the needs from different types of patient. By concerning the traditional risk factors, there is a positive correlation between psychological factors and metabolic syndrome as indicated in the research. In this research, it is hypothesized that the users of blood glucose monitoring mobile system tend to have different aspects for acceptance of healthcare technologies from the normal subject. In terms of previous literature, we analyzed several adoption areas such as new technology, psychological and personal and identified the key factors for user to accept the blood glucose monitoring mobile system. Through a survey process, 462 valid questionnaires were collected and processed via SPSS 20.0 to have description statistics and analysis of reliability and validity. Smart PLS 3.2 was used for examination of research model and verification of explanation power. From the results, the users who accept new technology, the performance expectancy, effort expectancy, social influence, facilitating conditions and habit have impact on users’ acceptance of blood glucose monitoring mobile system. From psychological aspect, users tend to be influenced by relative advantages, but not reference price and reference situation for blood glucose monitoring mobile system acceptance. From personal aspect, self-efficacy, trust, risk preference, health consciousness will affect the users to accept the system. Looking into the users’ behavior, facilitating conditions, habit and behavioral intentions have impact on the acceptance of the system from the users. According to the analysis results, the key factors are verified and explained in terms of influencing the users to accept the system. We hope the results will make contributions to the academic field and industrial companies
利用Burrows-Wheeler轉換對社會網路進行壓縮;Burrows-Wheeler Transformation and Social Network Compression
No full text[[abstract]]資料壓縮可有效將資料量減少。Burrows和Wheeler提出了一個轉換方式使轉換後的資料可以更有效的被壓縮。隨著社會網路的快速成長，如何將這些巨量資料進行無失真的壓縮成為一個重要的議題。在這篇研究中，我們提出兩個對社會網路進行無失真壓縮的演算法。我們的方法是將游程編碼做一些變化，並與Burrows-Wheeler轉換和霍夫曼編碼結合，對社會網路進行壓縮。
Data compression is an useful technique to reduce the size of data. Bur-rows and Wheeler proposed an algorithm which output can be used as aninput for another compression algorithm to achieve better compression ratio.With the rapidly expansion of social networks, how to save these enormousdata without losing any information is an important issue. Therefore, in thisthesis we proposed two methods to compress social networks. Our methodmodifies run-length encoding and combines Burrows-Wheeler transform andHuffman encoding to compress social networks
整合三軸線上學習控制演算法與LinuxCNC之實作探討;An Implementation Study of Integrating Three-Axis On-Line Iterative Learning Control Algorithm into LinuxCNC
No full text[[abstract]]本論文實作出嶄新的學習控制執行方法 --- 線上學習控制演算法。傳統的離線式學習控制演算法必須在機器加工後才能開始計算新命令，新型的線上學習控制演算法則可以在加工期間，同時進行新命令的計算，降低加工間的等待時間，提升量產效率。 本研究可分為三大工作。第一，將傳統離線式學習控制演算法進行大幅度優化，縮短計算單個命令所需的時間，經過優化後，提升8.06倍效能。第二，將離線式學習控制演算法改寫為線上學習控制演算法。由於新型演算法更加節省記憶體的特性，程式所需的記憶體減少99.2\%。第三，將線上學習控制演算法整合至開源開放CNC軟體 --- LinuxCNC。在LinuxCNC的控制介面中新增學習控制功能，讓使用者易於操作線上學習控制演算法。本研究將研究成果整合至五軸工具機中，實際測試並驗證演算法的可行性。實驗證實，線上學習控制演算法確實可在加工的同時進行學習控制運算，運算後的新命令也確實有效地降低加工誤差。
In this thesis, we formulate an innovative approach to realizing Iterating Learning Control (ILC) algorithm, on-line ILC. Traditional off-line ILC must be executed after the machining process of machine tools. The proposed on-line ILC could be executed simultaneously during the machining process, which reduces the idle time between processes, raises mass production efficiency. The thesis consists of three works. First, we significantly improved the performance of off-line ILC. The optimized version can achieve 8.06 times speedup. Second, we modified off-line ILC to on-line ILC. The on-line ILC can reduce the memory footprints by 99.2\% compared to the off-line ILC version. Third, we integrate on-line ILC into an open-source CNC software, LinuxCNC. A new GUI specific to on-line ILC is developed to make users easy to operate ILC function. The whole CNC system is evaluated on a five-axis machine tool. For the tested G-code, the experimental results show that on-line ILC can correctly run during the machining process of machine tools. The actual contour errors are reduced after machining with the commands computed by the on-line ILC
基於Bit-plane分解與重組應用於移動物體偵測;Bit-plane decomposition and restructuring for moving object detection
No full text[[abstract]]近年來，視頻監控越來越受歡迎，視頻監控技術的發展已不可或缺，因為它可以降低人工成本和檢測錯誤率。大多數視頻監控攝像機是固定的，因此對背景建模的發展需求增加。當運動對像出現在視頻中時，運動對像被稱為前景對象，其可以通過“背景減除”來提取，在視頻監控中，移動物體檢測已經成為自動了解視頻內容的核心技術之一。儘管已經提出了許多檢測方法，但是它們仍然具有一些問題，諸如運動對象的陰影和不完整形狀。在本論文中，我們對於視頻監控系統中的運動物體對象偵測，利用Bit-planes方法，進行了色彩資訊及YUV色彩空間資訊方面生成hyper bit-plane，而hyper bit-plane利用了YUV色彩空間的資訊抑制了陰影生長，而我們所的方法利用hyper bit-plane作為指標影像，而為了能夠清楚描繪出移動物體的輪廓及完整的移動物體，利用灰階Bit-plane影像作為較低基準的影像資訊，再將hyper bit-plane和低基準的影像資訊結果，利用時間域、空間域的特性，應用於圖像遲滯閾值和運動補償來補償空間和時間信息的損失，經由遲滯閾值和運動補償能夠將移動物體的輪廓及物體作整合。最後、我們將所提出的技術能夠在運用在即時監視系統上所需的效率，以提升視訊監視系統能夠快速做出系統所需要的應用。
In recent years, video surveillance has become increasingly popular, the development of video surveillance technology is indispensable, because it can reduce labor costs and detection error rate. Most video surveillance cameras are fixed, so there is an increased demand for background modeling. When the motion object appears in the video, the moving object is called the foreground object, which can be extracted by "background subtraction". In the video monitoring, moving object detection has become one of the core technologies of automatic understanding of video content. Although a number of detection methods have been proposed, they still have some problems, such as shadows and incomplete shapes of moving objects. In this thesis, we use the Bit-planes method for the motion object detection in the video surveillance system to generate the hyper bit-plane by the color information and YUV color space information. Hyper bit-plane uses the YUV color space information to suppress the shadow growth, and our method uses hyper bit-plane as the indicator image. In order to describe clearly the complete contours of moving objects, the use of grayscale Bit-plane image as a lower benchmark image information, and use spatial and temporal information to integrate hyper bit-plane and low Image information results, that applied to the image hysteresis threshold and motion compensation to compensate for the loss of space and time information, that can obtain complete foreground through the hysteresis threshold and motion compensation. Finally, we will be able to use the method in the real-time surveillance system
基於序列到序列迴遞神經網路之文字摘要應用於閱讀分析;Text Summarization Based on Sequence-to-Sequence with RNNs for Reading Analysis
No full text[[abstract]]序列到序列學習是一種自然語言處理的新技術，它是一種框架能讀取變動長度的輸入文字與產出變動長度的輸出文字，最大的特點是輸入輸出的長度不需對齊。從抽象來看，這種框架包含主要兩個主成分，一個是編碼器負責讀取文字並壓縮到語意空間，另一個是解碼器負責根據該語意產生輸出文字。如果將該框架運用在自動文字摘要，其輸入就是一篇文章，輸出則是一篇摘要。 本篇論文著重於探討序列到序列學習其技術原理，並將之應用在自動文字摘要的任務上。與一般分類任務不同，自動文摘的任務必須要在可以理解語意的前提下才能產生出好的文摘。在本篇的實作方法上並非擷取原文來產生文摘，而是根據字典逐字產生文摘，況且自然語言的字詞是數以萬計的，這些因素使得在計算上非常複雜。 實驗的訓練數據顯示，序列到序列學習應用在自動文摘是可行的；而驗證數據表示該系統還有持續研究的空間。在未來，除了持續研究該系統外，還可以發展藉由文摘的技術輔助文字分析的系統。
The sequence-to-sequence learning is a novel technique for natural language processing. It is a type of framework that can read an input sequence with variable length and generate an output sequence with variable length. The feature of this framework is the length of the input and output without alignment. Briefly, the framework contains two of main components. One is the encoder which is responsible for reading words and compressing into a linguistic space. The other one is the decoder which is responsible for generate words conditioned on the linguistic representation. If the framework is used in the automatic text summarization, the input will be an article, and the output will be a summary. This study focuses on the technical principles on the sequence-to-sequence learning, and the technique is used in the automatic text summarization. Unlike the general classification task, the task of automatic text summarization produces a suitable summary conditioned on more previous information. The implementation of method is not to extract the original article to produce a summary. The model generates words according to its vocabulary. Moreover, the natural language words or phrases are tens of thousands. Because of those factors, the computation and implementation are very complex
針對低覆蓋率定序區域之基因體重組軟體;Genome Reassembler for Low-coverage Sequencing Regions
No full text[[abstract]]第三代定序(TGS)可以在一個更短的時間內定序出更長的序列，而這將會成為序列組裝的首選，不幸的是，大多數大型基因定序的資料都無法滿足現有軟體所需要的定序深度，即使達到期望的定序深度，整個序列的覆蓋率仍然不均勻。本論文設計並實作了一個基因體重組器，針對在低覆蓋率區域中，重新判斷序列的重疊與組裝，並開發了新的降維技術(dimensional-reduction)在低覆蓋率區域有效的計算重疊。在實驗結果中，我們的方法在低覆蓋率(≦30x)的資料下改進了Canu、Falcon以及Miniasm的組裝結果，對於大型基因體中，覆蓋率從低到高中也可以看出結果的改善。
Third-generation sequencing (TGS) can produce much longer reads within shorter turnaround time, which is becoming the preferred choice for de novo genome assembly. Unfortunately, most large-genome sequencing projects are not affordable of sufficient sequencing depth required by existing TGS assemblers. Even the desired sequencing depth is achieved, the coverage across the entire genome is still uneven. This thesis designs and implements a genome re-assembler which aims to improve existing TGS assemblers by recovering missing overlap and performing re-assembly in low-coverage regions. A novel dimensional-reduction technique is developed for efficient overlap computation tailored for low-coverage regions. Experimental results indicated that our method improves assembly of Canu, Falcon, and Miniasm under low-sequencing depth ( 30x). For large genomes, the results of low- and high-sequencing depth are both improved probably owing to larger coverage variance
以空間、時間、轉換、及時空特徵作無參考視訊品質評估;No-reference Video Quality Assessment Using Spatial, Temporal, Transform, and Spatiotemporal Features
No full text[[abstract]]隨著多媒體傳輸的快速發展，人們對於視訊品質的感知愈來愈重視。受限於網路傳輸中有限的頻寬，無參考視訊品質評估技術是最具有彈性的方法。本研究是基於無參考並且擷取空間、時間、轉換及時空特徵預測視訊品質分數。首先，主要擷取邊緣偵測和塊狀資訊當作空間特徵。接著，亮度和運動的變化被視為是時間特徵。此外，離散餘弦轉換和小波轉換被用來擷取轉換特徵。有鑑於鮮少同時考慮空間和時間資訊，軌跡的統計特性和三維離散餘弦轉換被擷取當作時空特徵。然後，採用支持向量回歸預測視訊品質分數。根據LIVE video quality assessment資料庫進行預測的實驗結果顯示本研究明顯優於其他現有的方法。
With the rapid development of multimedia transmission, human attach more importance to perceptual video quality. In terms of finite bandwidth in the network transmission, no-reference video quality assessment is more flexible to estimate video quality. The proposed approach based on no-reference utilizes spatial, temporal, transform, and spatiotemporal features to predict video quality score. Above all, edge detection and blockiness are extracted in spatial domain. Then, intensity and motion variations are considered as temporal features. In addition, discrete cosine transform and wavelet transform are utilized to extract features in transform domain. In view of the fact that spatial and temporal information is rarely taken into account simultaneously, statistical characteristics of trajectory and three-dimensional discrete cosine transform are extracted as spatiotemporal domain features. Then, support vector regression is applied to predict video quality score. Based on the experiment results on LIVE video quality assessment database, the performance of the proposed approach is significantly better than other existing approaches