This course introduces fundamental problems in computer vision and solutions to them. For most of the topics, classical methods and their deep learning counterparts will be presented together. As this is an introductory course, more emphasis will be given on practical solutions rather than theoretical ones.

Students are expected to have basic math knowledge such as linear algebra, probability theory, and optimization. Programming skills are also required. In particular, students are expected to be familiar with deep learning libraries and Linux environment for their final projects, there is no restriction on specific languages and computing environments in the course though. Please note that this course will not provide programming tutorials as its targets are graduate and senior undergrad students.

Midterm exam (40%) / Final project (60%)
The percentages are subject to change.

There is no regular textbook, but the following will be useful as supplementary materials.
1. Computer Vision: Algorithms and Applications, Richard Szeliski. (Online version is available at POSTECH library.)
2. Computer Vision, A Modern Approach, Forsyth, Ponce, Pearson Education Inc.

- Course Overview
- Preliminary: Probability and Linear Algebra
- Image Representation and Classification
- Object Detection
- Semantic Segmentation
- Metric Learning and Image Retrieval
- Video Representation and Classification
- Object Tracking
- Matching and Fitting
- Camera Models
- 3D Geometry

This course will be given in an offline manner (501, Cheong-am Library), but a few classes could be given in forms of pre-recorded videos with prior notices.