Repository for the Visual Computing course (VCO2-IM.ma VZ SS24) at the FH Hagenberg.
Contact: David C. Schedl.
| # | Tutorial (link to .ipynb) |
Open in Colab |
|---|---|---|
| 1 | Python Tutorial |  |
| 2 | OpenCV Tutorial | |
| 3 | Filters | |
| 4 | Edges and Lines | |
| 5 | Corners and Featuers | |
| 6 | Alignment | |
| 7 | Stereo | |
| 8 | (recap) Neural Networks | |
| 9a | CNN (LeNet in TensorFlow) | |
| 9b | CNN (LeNet in PyTorch) | |
| 10 | Transfer Learning a CNN | |
| 11 | Image representation with an MLP | |
| # | Homework (link to .ipynb) |
Open in Colab |
|---|---|---|
| 1 | Compression | |
| 2 | ML Compression | |
Students have the option to run the code online with Google Colab (requires a Google account) or locally with a local installation of Python.
Everything runs on a Google machine, so you don't need to set up anything on your computer. Furthermore, the machines come with the most popular libraries preinstalled.
Just click on the corresponding Open in Colab badge:
Install Python on your computer via Conda/Miniconda or the Python Installer. Use Python3, as Python2 is not supported anymore. Furthermore, you need an Editor that supports Jupyter (.ipynb) notebooks. I recommend using Visual Studio Code. Optionally, you can also use a local server and open Notebooks in your browser (Visual Studio simplifies this).
- Python Documentation
- OpenCV Tutorial
- If you know Matlab, you can find the differences between Matlab and Python here.
This course will be graded based on your performance in the course homeworks. The homework tasks will be announced while we progress through the course.
Furthermore, you will give an oral presentation about a related computer vision field. The topic of your presentation is free to choose (as long as it is CV related). Focus on a Computer-Vision task/algorithm (a scientific work), understand its details (advantages, limitations, competition, ...), and explain it to your fellow students.
IMPORTANT: Pick a fun topic that interests you! You can find further details here and an inspiration for topics here.