Course Description
This course covers several fundamental and interesting topics in computational photography and computer
vision, such as colorization, texture synthesis, inpainting, HDR imaging, tone mapping, deblurring,
stabilization, super-resolution, matting, warping, matchmoving, and so on. The goal of this course is to learn
the selected topics and know how they can be used to create visual effects in images and videos. Students
are expected to get hands-on experience of related techniques.

Computer Vision for Visual Effects, Richard J. Radke, Cambridge University Press,

Computer Vision: Algorithms and Applications, Richard Szeliski, 2010

Teaching Method
3-hour lectures and discussions every week
All assignments, readings, and projects are done in groups.
Each group contains 2-3 students.

1 introduction, image formation, color, color harmonization, color grading
2 colorization, chrominance blending
3 texture synthesis, image analogies
4 inpainting, seam carving, internet-based inpainting, sparse coding
5 HDR imaging, tone mapping
6 bilateral filtering, nonlocal means, joint bilateral filtering
7 gradient domain compression, HDR related
8 intrinsic images, relighting
9 graphical models
10 matting
11 poisson image editing, patch-based editing
12 warping and morphing, rigid morphing
13 faces
14 features and matching
15 matching and dense correspondences
16 dense correspondences
17 matchmoving
18 term project poster

80%: 10 assignment + 4 projects
10%: weekly readings
10%: participation

Course website