5516 - 電腦視覺實務與深度學習
Computer Vision Practice with Deep Learning
教育目標 Course Target
電腦視覺已經深入生活當中，然而在不同應用情境下存在不同的實務挑戰，本課程講授學術文獻上以深度學習為主流之最新電腦視覺技術解決方案，以期建立修課學生之電腦視覺學理基礎與實務應用能力。
Computer vision has been deeply embedded in daily life, but there are different practical challenges in different application scenarios. This course teaches the latest computer vision technology solutions with deep learning as the mainstream in academic literature, with a view to establishing the computer vision theoretical foundation and practical application capabilities of the students.
參考書目 Reference Books
．The primary teaching materials will consist of teacher-created slides.
．No textbook but recommended readings:
． The primary teaching materials will consist of teacher-created slides.
． No textbook but recommended readings:
評分方式 Grading
| 評分項目 Grading Method |
配分比例 Percentage |
說明 Description |
|---|---|---|
|
Assignments Assignments |
50 | (tentative) 3 times of programming assignments, announced in class or from the course website |
|
Class Participation Class Participation |
10 | Only applicable for “Invited Talks”. Rules for check-in/check-out: ．Can get 10 points if check-in before the talks (9:10am) and check-out after the talks (12:10pm) ．Every 10 minutes short will get -1 point penalty (e.g., check-in during 9:10am-9:20am and check-out after the talks) |
|
Final Projects Final Projects |
40 | Team-based presentation with a proof-of-concept implementation of a selected ‘top-level’ and ‘new’ (published from 2025 or after) paper, which must be enhanced by the team’s own innovations |
授課大綱 Course Plan
點擊下方連結查看詳細授課大綱
Click the link below to view the detailed course plan
相似課程 Related Courses
無相似課程 No related courses found
課程資訊 Course Information
基本資料 Basic Information
- 課程代碼 Course Code: 5516
- 學分 Credit: 3-0
-
上課時間 Course Time:Friday/2,3,4[遠距課程]
-
授課教師 Teacher:鄭文皇/陳倫奇
-
修課班級 Class:共選修3,4,碩博1,2
-
選課備註 Memo:教育部補助臺灣大專院校人工智慧學程聯盟，開設學校：臺灣大學（鄭文皇），上課時間：星期五 9:10-12:10，開放大三以上修習，遠距課程
交換生/外籍生選課登記
請點選上方按鈕加入登記清單，再等候任課教師審核。
Add this class to your wishlist by clicking the button above.