o-net technologies的問題，透過圖書和論文來找解法和答案更準確安心。 我們挖掘到下列精選懶人包

A Study of Competency Model for a Project Portfolio Manager: Triangulation Approach

為了解決o-net technologies的問題，作者廖琉沙 這樣論述：

Among various tools to track the implementation of changes and innovations, one of the top places is occupied by project portfolio management (PPM). With the enhancing importance of PPM, the attention to the person actively manages this process – a portfolio manager (PfM) – is increasing respective

ly. Organizations strive to have competent PfMs. Although the role is formally defined in professional standards, there is still a lack of description and ambiguity in needed PfM competencies. This research was aimed to offer the competency model of PfM by triangulating three sources: literature, jo

b descriptions, and opinions of the professionals in the PPM area. The framework of the study was built according to the purpose and involved such methods as content analysis and corpus analysis. The research findings reflected the primary purpose: were analyzed corpora of competency-related textual

information, each corpus was checked on the representativeness and used in topic modelling. The experts verified the international intersectoral competency model for PfM. The proposed competency model for PfM contains six competencies: business focus, diverse expertise, expedient communications, pr

ocess stewardship, self-management, and transformational demeanour. Each element contains six to eight indicators divided into two levels – critical and desirable. Hence, the introduced competency model could be used to clarify PfM roles in the organisations and for the individual professional devel

opment of the PfMs.

應用人臉辨識技術之課堂點名系統

為了解決o-net technologies的問題，作者陳充鑫 這樣論述：

隨著科技日新月異，為了提高課堂中的教學效率而應用了許多資訊技術，例如以往的點名都是由老師唱名學生應答來做出席依據，後來新增了許多不同的點名方式，有的會用學生證作為媒介，讓學生透過感應的方式記錄出缺席，有的會用學生系統點名，學生必須使用學號加密碼登入自己的帳號才能夠點名，也有透過簽到的方式做點名紀錄，學生必須親自簽名才算數，但這些方法都有各自的弊端存在，學生證感應的方式是認卡不認人，就會有代替別人刷卡的作弊方式出現；學生系統點名則是可以將帳密交給別人代替登入點名；而學生親自簽名雖然可以避免別人冒名代簽的狀況發生，但跟唱名的方式點名一樣耗時耗力。為了方便學生和老師的使用及減少課堂點名的缺點，本論

文規劃運用人臉辨識技術應用於課堂點名系統，既可以降低課堂點名的弊端發生，也可以省下以往透過唱名的點名方式所耗費的時間與精力。該點名系統會利用架設在教室高處的相機鏡頭拍攝課堂上所有的學生，經由MTCNN演算法找出照片中所有學生的臉，最後由FaceNet模型辨識出有哪些學生在此照片中，藉此達到人臉辨識點名的效果。本論文製作過程中，設計了一個系統整合UI介面，整合多執行緒即時影像擷取及顯示模組(IP Cam/Web Cam)、課程分類、學生管理、深度學習人臉偵測技術MTCNN與深度學習人臉辨識技術FaceNet。為了驗證本論文整合系統的可行性，我們會收集至少80位學生的人臉影像，作為實際使用的測試。

所提出之整合系統使用中低階圖形處理器(GTX1660Ti)即可達到97.3%的準確率以及20~25FPS的即時運算。