影像編修技術與特效合成
Image Manipulation Techniques and Visual Effects
| 節 | 週四 |
|---|---|
5 13:20–14:10 | 影像編修技術與特效合成 EC115（光復） 2 節連堂 |
6 14:20–15:10 |
* 根據陽明交大上課時間表所列
Digital manipulation techniques are now popularly used to generate compelling, plausible, or exaggerative images for photo retouching or visual effects. In this course, we will introduce modern image analysis and synthesis techniques, e.g. image matting, segmentation, warping, compositing, inpainting, style transfer, and so forth. Students will learn the algorithms behind, such as objective optimization and deep learning. Students will develop intelligent editing tools for their projects.
C/C++ or Python programming, calculus, matrix computation or linear algebra (Optional) {Backgrounds in computer graphics, computer vision, image processing or deep learning will be helpful }
無備註
Course-related material will be uploaded onto E3. {Class lectures and the teaching materials (slides, homework, etc.) are in English. Students are encouraged to speak English for questions and discussion, but we are still welcome to use Mandarin Chinese to clarify certain issues.}
(Provisional) #Programming homework assignments (40~60%) #Paper study and presentation (10~20%) #Term project (35~50%) #Class participation (0~10% or bonus)
教師未提供此項資料
| 週次 | 主題 |
|---|---|
| 第 1 週 | Overview of image manipulation and visual effects |
| 第 2 週 | Essential principles of images and features. Color histograms and GMM |
| 第 3 週 | Image matting and compositing |
| 第 4 週 | Image segmentation and compositing with Graphcut |
| 第 5 週 | Image warping and morphing |
| 第 6 週 | Texture manipulation |
| 第 7 週 | Patch-based optimization and image inpainting |
| 第 8 週 | Essential concepts about neural networks for images |
| 第 9 週 | Convolutional Neural Network (CNN) and Generative Adversarial Network (GAN) |
| 第 10 週 | Image manipulation and visual effects with CNN, GAN and recent models (I) |
| 第 11 週 | Image manipulation and visual effects with CNN, GAN and recent models (II) |
| 第 12 週 | Image manipulation and visual effects with CNN, GAN and recent models (III) |
| 第 13 週 | Advanced topics |
| 第 14 週 | Paper survey and project proposal (I) |
| 第 15 週 | Paper survey and project proposal (II) |
| 第 16 週 | Week for alternative curriculum/supplementary teaching |
| 第 17 週 | Week for alternative curriculum/supplementary teaching |
| 第 18 週 | Final presentation and demo |
Course slides References: *Proceedings of ACM SIGGRAPH, SIGGRAPH Asia, IEEE CVPR, ICCV, ECCV, and other related articles.
- 地點
- EC704
- 時間
- To be announced.
- 聯絡方式
- Email: ichenlin@cs.nctu.edu.tw