Introduction

Key Takeaways:

The course is about perception for self driving cars using computer vision and deep learning.
Perception and Computer Vision form about 80% of the work that Self-Driving Cars do to drive around.
The course will cover different tasks that a Self-Driving Car Perception unit would be required to do, such as:
- Road Segmentation: identifying the drivable area of the road using a Fully Convolutional Network (FCN).
- 2D Object Detection: locating and classifying objects in the image using You Only Look Once (YOLO) algorithm.
- Object Tracking: tracking the movement and location of objects over time using Deep SORT algorithm.
- 3D Data Visualization: transforming and projecting 3D data from LiDAR sensors using Homogenous Transformations.
- Multi Task Learning: performing multiple tasks simultaneously using a Multi Task Attention Network (MTAN).
- 3D Object Detection: detecting and localizing objects in 3D space using SFA 3D algorithm.
- Camera to Bird's Eye View: converting the camera image to a top-down view of the scene using UNetXST algorithm.

All the datasets and notebooks code are provided for each project.

KITTI dataset: real-world data, which is a benchmark for autonomous driving research.

Perception for Self-Driving Cars — Free Deep Learning Course.
Computer Vision and Perception for Self-Driving Cars (Deep Learning course) - YT/@freecodecamp.
Computer Vision and Perception for Self-Driving Cars (Deep Learning Course) - classcentral.com.