1) Behavior Cloning - Using Keras to train a neural network to drive a car using training data from manual driving (Result)
2) Traffic Sign Classifier - Using Tensorflow (LeNet-5) to classify traffic signs (test accuracy 92.6% compared to AlexNet 85% accuracy)
3) Capstone Project - Written in Python using ROS. Navigates a simulated car through the track and detects traffic lights as necessary using DarkFlow.
4) Lane Detection (Method 1) - Uses OpenCV and the canny edge detection method to detect the lanes. (Result)
5) Lane Detection (Method 2) - Uses OpenCV to calibrate the camera with checker cards and projects the video to bird-eye view and detects the lane with a special line algorithm. (Result)