We study the creation of physical adversarial examples, which are robust to real-world transformations, using a limited number of queries to the target black-box neural networks. We observe that robust models tend to be especially susceptible to foreground manipulations, which motivates our novel Foreground attack. We demonstrate that gradient priors are a useful signal for black-box attacks and therefore introduce an improved version of the popular SimBA. We also propose an algorithm for transferable attacks that selects the most similar surrogates to the target model. Our black-box attacks outperform state-of-the-art approaches they are based on and support our belief that the concept of model similarity could be leveraged to build strong attacks in a limited-information setting.

近年來原住民語使用逐漸減少，面臨傳承的危機。本研究製作太魯閣族語音評分系統，希望幫助人們學習太魯閣族語。利用原委會網站的音檔和田野調查收集錄製的族語語音與族語老師的評分為本研究的研究語料，以Kaldi為語音辨識框架來製作語音辨識模型。對模型輸出的音素與正確單詞進行字串相似度計算，將計算數據與模型產出的發音品質分數作為選擇模型的依據，透過機器學習方法選擇最接近族語老師評分的模型。以此模型計算研究語料，再以機器學習製作一個七成正確率的評分分類器，最後實作出評分系統網站。此網站除了可以學習族語之外，也可以幫助未來繼續收集更多語料，提高評分系統的正確率。期望幫助人們學習太魯閣族語，讓原住民族語原音得以重現。

在資訊爆炸的時代，效率閱讀、整理資料的能力越趨重要。身為高中生，學習時的閱讀量龐大，還須另外自己挑選重點句，重新整理筆記。因此我想如果可以讓電腦自動摘取文章的重點，就能幫助學生效率學習。 大多數現存的自動摘要研究適用於英文文本，本研究利用演算法抓取中文文章的摘要，使學生可以真正實用該演算法於日常學習當中。除此之外，此研究比較了不同方法摘要的準確率以及優缺點。

本研究旨在發展一套可限制特徵之人臉生成器，能依據目擊者對嫌犯的特徵描述，使用真實人臉照片進行特徵生成，以生成嫌犯臉孔。本研究以生成對抗網路為中心，配合選擇性傳輸單元、編碼器與解碼器結構，發展出五官、毛髮、飾品、整體形象四大類別，共有40個特徵生成編輯種類。搭配友善的圖形使用者介面，讓使用者可將已相關度篩選後之真實人臉資料庫或是另行輸入之照片，因應不同的嫌犯臉孔描繪需求情境，對該圖片臉孔進行指定特徵之生成編輯。此技術除了可輔助刑案目擊者回憶嫌犯臉孔特徵，亦可應用於眼鏡業、醫學美容、美髮業等領域。

到遊樂場玩，看到刺激的海盜船，想探討它對心臟的影響。由於血壓無法在短時間測得，所以就由血流速率的變化，推測血壓的變化。接著就製做一台心臟模擬器，再裝上攝影機，模擬各種可能的情況，再由電腦播放，找出各種變因的影響。 由坐姿的研究發現，向上流的血液，在最低點時，受到離心力向下的影響，使血液流速減緩，但是卻使向下流的血液流速增加。再比較海盜船的擺長影響，發現比較長的擺長，在坐姿時，向上流的血液受到的離心力較大，流速減緩的較多。在不同姿勢的比較實驗，坐姿的血液流速變化百分比較小，臥姿的變化非常大。 坐海盜船之前要先測量血壓狀況，評估是否有心血管疾病，再選擇適當時間搭乘，才是神鬼奇航的安全通道。

In Singapore’s evolving cyber landscape, 96% of organisations have suffered at least one cyber attack and 95% of organisations have been reporting more sophisticated attacks in the frame of one year according to a 2019 report[1] by Carbon Black. As such, more tools must be utilised to counter increasingly refined attacks performed by malicious actors. Honeypots are effective tools for studying and mitigating these attacks. They work as decoy systems, typically deployed alongside real systems to capture and log the activities of the attacker. These systems are useful as they can actively detect potential attacks, help cybersecurity specialists study an attacker’s tactics and even misdirect attackers from their intended targets. Honeypots can be classified into two main categories: 1. Low-interaction honeypots merely emulate network services and internet protocols, allowing for limited interaction with the attacker. 2. High-interaction honeypots emulate operating systems, allowing for much more interaction with the attacker. Although honeypots are powerful tools, its value diminishes when its true identity is uncovered by attackers. This is especially so with attackers becoming more skilled through system fingerprinting or analysing network traffic from targets and hence, hindering honeypots from capturing more experienced attackers. While substantial research has been done to defend against system fingerprinting scans (see 1.1 Related Work), not much has been done to defend against network traffic analysis. As pointed out by Symantec[2][3], when attackers attempt to sniff network traffic of the system in question, the lack of network traffic raises a red flag, increasing the likelihood of the honeypot’s true identity being discovered. In addition, the main concern with regards to honeypot deployment being their ability to attract and engage attackers for a substantial period of time, an increased ability to interest malicious actors is invaluable. Producing human-like network activity on a honeypot would appeal to more malicious actors. Hence, this research aims to build an intelligent web-surfer which can learn and thus simulate human web-surfing behaviour, creating evidence of human network activities to disguise the identity of honeypots as production systems and luring in more attackers interested in packet sniffing for malicious purposes.

本研究以AI虛擬試衣系統(Virtual Try-on)為主題，透過深度學習技術，並結合幾何匹配模型，開發出試衣系統，可將使用者上傳的照片，模擬成穿著新衣的模樣。 首先，以深度學習模型將人物原始圖片取出骨架節點，並生成人體遮罩以及保留人物頭部，再結合以上三種資訊合成為高維特徵圖。接著將目標替換衣物生成出依照人體姿態扭曲後的衣物圖片。最後於Virtual Try-on模型中將人體高維特徵圖與扭曲衣物作為輸入，並經過深度學習網路合成出穿著目標衣物之人體圖像。本研究結果發現，人物站姿單純，且雙手緊貼身側，以及拍攝角度為正面、衣服款式為短袖、背景色彩對比度較高與衣服圖案單純的原始圖片，可得到較好的合成結果。

Nowadays, all blind people are able to read any type of text thanks to The Braille system, which is a system of reading by touch. The Braille system provides a good and important help to blind people but it’s not enough. They try to integrate into society, try to read and write the same as those blessed with sight. However, only 18% of people can actually write in Braille and doing so takes a lot of effort. While Braille techniques have existed for many years now, there isn’t one that takes advantage of the comfort and easiness of modern technology. To be fair and give more chances to the blind people, and with all the available technologies nowadays, we must find solutions and innovative ideas to fulfill this objective. And this is how our project was born. To be honest, it’s not just a project, for us it’s a dream. Our focus in this project is creating an automatic machine based on transforming our language to their language and transforming this new invention to assist the blind in writing whole texts through speaking. It will be easier than the traditional way. This part of the project will help us print Braille on paper automatically in the cheapest, fastest and easiest way, unlike the traditional method that not only takes a lot of time but also needs a professional. Our printer will only cost approximately 250$ compared to regular printers that cost 6500$! the principles of the code of the speech recognition are voice pickup in all languages as first step, second speech recognition, then speech translation, after that speech analysis and finally translation into Braille and printing it. As for the future, we will work on face, object and image recognition as a scanner to make the visually impaired people read the name of the object and recognize people’s faces like anyone else.

做研究時，經常需要閱讀多篇長篇論文，如何快速理解並歸納重點是影響效率收益的關鍵任務。針對此問題本研究運用深度學習配合其他程式算法輔助。將整個問題拆解成三個主要部分：模型處理、結構剖析、心智圖輸出，模型處理的部分我們訓練了三種不同的Seq2seq模型實現文本摘要，分別使用LSTM、BERT、ALBERT、比較並選擇綜合效益最好的模型；而負責拆解論文的結構剖析系統，則利用論文目錄分解整份論文，最後，心智圖輸出系統則整合剖析系統與模型結果，再調用Xmind API去生成架構清晰易讀易理解的論文心智圖。

全球貧血人口普遍，然許多人並不了解自身是否罹患貧血；長期患有貧血的病人，亦需定期抽血檢驗追蹤是否有貧血惡化達到需接受輸血的程度。研究顯示，結膜之顏色與貧血有絕對關係，結膜越白則貧血越嚴重，醫師也常使用結膜顏色推測是否有貧血情形。若能設計手機軟體自動分割結膜影像並分析其顏色，將有機會推測受試者是否罹患貧血。本研究收集22位無貧血者及8位貧血病人，並獲得其近期血紅素數值。以手機取得受試者之眼睛影像後，成功設計程式以深度學習完成結膜自動影像分割，對於分割影像以面積大小進行後期處理後，依其取得下眼瞼結膜之三原色平均，再利用kNN與SVM演算法判斷預測出該受試者是否具有貧血之症狀。本研究主要分為兩階段，其一為進行下眼瞼結膜分割模型訓練；其二為製作有無貧血之判斷模型。整合上述眼瞼分割模型(IoU=89.8%±0.02%)與貧血判斷模型(SVM以polynomial核函數測出 準確值93.3%±24.3%)後，可得貧血診斷準確率為80%。此結果代表AI技術有機會透過結膜影像，判斷被拍攝者是否有貧血情形，未來若能增加研究人數，將可設計網頁版或手機APP加以推測血紅素值，供大眾居家篩檢。

This work is devoted to solving the problem of orientation in the space of visually impaired people. Working on the project, a new way of transmitting visual information through an acoustic channel was invented. In addition, was developed the device, which uses distance sensors to analyze the situation around a user. Thanks to the invented algorithm of transformation of the information about the position of the obstacle into the sound of a certain tone and intensity, this device allows the user to transmit subject-spatial information in real time. Currently, the device should use a facette locator made of 36 ultrasonic locators grouped in 12 sectors by the azimuth and 3 spatial cones by the angle. Data obtained in such a way is converted into its own note according to the following pattern : the angle of the place corresponds to octave, the azimuth corresponds to the note and the distance corresponds to the volume. The choice of the notes is not unambiguous. However, we used them for the reason that over the centuries, notes have had a felicitous way of layout on the frequency range and on the logarithmic scale. Therefore, the appearance of a new note in the total signal will not be muffled by a combination of other notes. Consequently, a blind person, moving around the room with the help of the tone and volume of the sound signals, will be able to assess the presence and location of all dangerous obstacles. After theoretical substantiation of the hypothesis and analysis of the available information, we started the production of prototypes of the devices that would implement the idea of transmitting information via the acoustic channel.

現今逐漸有許多科學家研究出一些能夠進行預測事件的方法，諸如天氣預報、價格變化的幅度等等。在眾多生活常見的問題中，我們選定違規停車為例，作為研究主題，而我們推測違規停車是一種具有週期性的問題，有多樣的潛在變因存在，於是我們跳脫以往多以心理層面、法規制度與地理熱點等的視角去看待違規停車，以資訊科學及數學的方法，建立一套分析的模式加以進行研究。 本文以連續小波變換CWT、最大重複離散小波轉換MODWT、差分整合移動平均自迴歸模型ARIMA、混合MODWT+ARIMA模型及混合Harr離散小波變換DWT+ARIMA模型，對臺南市三個不同特性的行政里的共五個聚類點，進行預測。原始的資料以週為單位，分成全日(allDay,0~24時)及白天(Day, 08~16時)兩種型態，並以連續46週的資料預測接下來4週。結果採用方均根誤差RMSE相對平均值、標準差、最大值的比值，進行預測品質的比較。