Module 4: Computer Vision from Foundations to Jetson Practice
This folder contains a rebuilt Module 4 designed as a systematic computer vision course for beginners.
Course Structure
Part I: Foundations and Theory
This part is suitable even for learners who do not yet have a Jetson device.
| Module | Topic | Type |
|---|---|---|
4.1 | Introduction to Computer Vision | Theory |
4.2 | How Computers Represent Images | Theory + OpenCV examples |
4.3 | Classical Computer Vision | Theory + OpenCV examples |
4.4 | Neural Networks and CNNs | Theory + simple code examples |
4.5 | Deep Learning Computer Vision Tasks | Theory |
4.6 | Train and Deploy Your Own Vision Model | Theory + code |
Part II: Edge Deployment and Jetson Practice
This part applies the earlier knowledge to edge AI deployment.
| Module | Topic | Type |
|---|---|---|
4.7 | Model Export and Edge Deployment | Theory + code |
4.8 | Real-Time Vision Pipeline Frameworks | Theory + code examples |
4.9 | DeepStream and Jetson | Theory + code examples |
4.10 | Frontier Vision Technologies and Outlook | Theory |
Appendix
This appendix is not part of the main 10-section spine. It is a practical project that extends the Jetson deployment half and gives learners a complete end-to-end example.
Recommended Baseline for the Jetson Practice Half
The deployment half of this course assumes:
JetPack 6.2.x- Jetson Linux
R36.4.x - CUDA
12.6 - TensorRT
10.3 - cuDNN
9.3 DeepStream 7.1Jetson Platform Services
Teaching Principles
This rebuilt Module 4 follows four teaching principles:
- Explain the "why" before the "how".
- Use code to illustrate concepts, not to replace explanation.
- Treat data, metrics, and error analysis as core topics.
- Keep deployment in the later half so the learner first builds understanding.
Suggested Learning Path
If you are a beginner, follow the sections in order from 4.1 to 4.10.
If you already understand computer vision basics and mainly want Jetson deployment, you can skim 4.1 to 4.6 and then focus on 4.7 to 4.10.