List of my submission for completion of Term-1 of Udacity's SDC (Self Driving Car) Nanodegree.

The aim of this project was using simple vision techniques to find the lane boundries in a video. This was done by finding HoughLines and then applying basic smoothing between frame-to-frame to track the road lines. More

Our first dive into the world of deep learning. In this project, using Convolution Neural Network (CNN), we implemented a traffic sign classifier trained on german traffic sign dataset. The project involved an extensive amount of experimentation along with learning about generating extra dataset from existing one using basic image manipulation, balancing dataset among classes and implementing different CNN using Tensorflow. More

The main aim of the behavioral cloning project was to train a convolution neural network to mimic human driving behavior using the open source Udacity Car Simulator. This document explains our approach to the problem along with the explanation of the training data and final model. More

Here we revisit our very first problem of finding road lanes, however, with a little more advance approach. Applying perspective transform on an undistorted image from the car camera, we obtain a bird-eye view of the road. Then using a combination of color and graident based thresholding along with a sliding window algorithm we determine the markers on the lane lines. Using a least-square polynomial fitting, we can compute the lane boundries including the curve and turns in the road. More

In this project, we implemented a vehicle detection and tracking algorithm using a combination of HOG, color and spatial filters along with a sliding window algorithm that uses a linear SVM classifier to detect vehicles. More