Hours:
18 hours (4 credits)

Room:

From remote by using Microsoft Teams. The link will be sent in due time to all students who registered to the seminar.

To register to the course, click here

Short Abstract:

This course introduces foundational representations, models and processes to extract 3D information from streams of images. It introduces the geometric camera model and the relations in a multiview scenario as well as the techniques to compute 3D structure from image sequences. In a second part 3D information is used in a class of problems involving 3D registration: Self-Localization and Mapping (SLAM), 3D shape recognition and tracking. Finally the course will analyse the recent trend towards data-driven tools based in deep-learning models. For each of these topics the a set of software tools will be introduced and the course will be evaluated.

Course Contents in brief:

1. Linear Algebra review
2. Perspective camera and transformations
3. 3D Reconstruction : BA and bilinear methods
4. 3D Registration : ICP
5. SLAM: Solving the Pose-graph
6. Applications and tools

Schedule:

1. Day1 - 19 April 2021 - 9.30-13.00
2. Day2 - 20 April 2021 - 9.30-13.00
3. Day3 - 21 April 2021 - 9.30-13.00
4. Day4 - 22 April 2021 - 9.30-13.00
5. Day5 - 23 April 2021 - 9.30-13.30