1) Overview

This repository demonstrates the latest technology being used right now in autonomous driving. The aim is to drive further progress and lead us to safer, more efficient, and widely accepted autonomous driving systems in the near future.

A variety of techniques are included in this project, such as

1. Road segmentation
2. 2D object detection
3. Object Tracking
4. 3D Data Visualization
5. Multi-Task Learning
6. Birds Eye View
7. 3D Object Detection

2) Goals

The goals of this project are to:

Learn about the different computer vision and perception techniques that are used in self-driving cars Develop and evaluate new algorithms for these tasks Contribute to the open-source community of self-driving car research

3) Methods

The methods that are used in this project include:

1. Deep learning
2. Traditional computer vision techniques
3. Data augmentation
4. Transfer learning

4) Results

The results of this project show that deep learning techniques can achieve state-of-the-art results on the tasks of lane detection, object detection, and traffic sign recognition. However, traditional computer vision techniques can still be effective, especially with limited data.

5) Challenges

The challenges that were encountered during this project include:

1. Limited data
2. Hardware limitations
3. The need for real-time performance

6) Future Work

The future work for this project includes:

1. Extending the system to new tasks, such as obstacle avoidance and path planning
2. Improving the accuracy of the algorithms
3. Making the system more robust to changes in the environment