Content

In this course, "Computer Vision" will primarily focus on the recognition of objects or events in images and videos. The lecture will be split into classical CV algorithms and modern solutions based on neural networks. 

1.Introduction

2.Edge Detectors

3.Histograms

4.Optic flow

5.Hough Transform

6.SIFT / SURF

7.Introduction to Neural Information Processing

8.Convolutional Neural Networks

9.Image Classification, Object Detection

10.Semantic Segmentation

11.Pose Estimation

12.Vision Transformers

13.Recurrent Neural Networks, Image Captioning

14.Generative Models

15.Unsupervised feature extraction

Lecture Format

We will have lectures on the topics outlined above. Those are either already available on YouTube or we will record and upload new ones (see also below on  How To Introduce New Topics). 

Lecture Videos

About 80% of the lectures will be from this list: 

https://www.youtube.com/playlist?list=PLs7Vp-pCDX7yrUmgkxAEdNcgriOU6IBg5 

For the rest we will create a new list, URL will be posted here. 

Lecture PDFs and Slides

Can be found in the resources section. If you have found typos, mistakes or are having a hard time understanding the contents, just let us know!

How To Introduce New Topics

If you think you have found a great paper, or have always been fascinated by a specific topic within the field of computer vision, please send us your suggestions via Mattermost (see below). We will have a vote to see how your peers think about this new topic. 

Mattermost

My research lab uses Mattermost, an open source messaging app. I would like to try it this semester for the course.  If you already have an account for the university hosted gitlab, you should be able to access the channel: https://mattermost.imp.fu-berlin.de/biorobotics/channels/ws-2425-computer-vision

If you are having problems logging in or accessing Mattermost, let me know.