Programming a Real Self-Driving Car
Capstone project for Udacity Self Driving Car Nano Degree
Developed using ROS and tested on Udacity's Self Driving Car, Carla
Implemented nodes for traffic light detection, waypoint following and control
Used a Fully Convolutional Network to label the pixels of a road in images.
Designed a path planner to create smooth, safe paths for the car to follow along a 3 lane highway with traffic
Model Predictive Control
Implemented a Model Predictive Control using a kinematic model to drive the car around the track.
Fitted the MPC by looking N points ahead spaced by dt seconds and fitting it to a third order polynomial.
Implement a PID controller in C++ to maneuver the vehicle around the track!
Utilized the cross track error (CTE) and the velocity (mph) from the simulator to compute the appropriate steering angle.
Implemented a 2 dimensional particle filter in C++.
Used the parameters from the simulator to accurately localize the vehicle position and yaw
Unscented Kalman Filter
Utilized an Unscented Kalman Filter to estimate the state of a moving object of interest with noisy lidar and radar measurements
Code written in C++
Extended Kalman Filter
Utilized a kalman filter to estimate the state of a moving object of interest with noisy lidar and radar measurements.
Code written in C++
Vehicle Detection and Tracking
Identified vehicles in a video from a front-facing camera on a car using HOG feature extraction and a Linear SVM classifier
Advanced Lane Finding
Used camera calibrationm, distortion correction , color transforms, gradients, etc., to detect and fit lane pixels
Used the image data and steering angles from the Udacity simulator to train a convolutional neural network (Keras) and used this model to drive a car autonomously around the track.
Traffic Sign Classifier
Classified traffic signs using deep neural networks and convolutional neural networks.
Used tensor flow for implementation
Finding Lane Lines
Identified lane lines on the road in an image and a video stream using python
Built a text summarizer to summazrize emails and articles
Used snowball stemmer, tokenizer and wordnet to filter
Used Naive Bayes for scoring
Robotic Arm Training
Used edge and blob detection to identify a food particle and trained a robotic arm using reinforcement learning to feed it to a mouth
Real time detection and tracking of a moving coloured object
Developed algorithms to trace the path of a colored moving objects. Programmed algorithms in MATLAB and Python(OpenCv)
Sign Language Identifier for the Visually Impaired
Used Principal Component Analysis for feature extraction and k-Nearest Neighbors to classify Sign Language alphabets.(MATLAB)
Used Gesture Dataset (2012) provided by Massey University for training and testing.
Soil Nutrient Analysis
Designed a portable embedded instrumentation systems (Arduino nano) to analyze the nutrients contents in soil using spectroscopic methods.
Designed and built android app to display results over Bluetooth.
Dipole Antenna Analysis
Analyzed the various properties of thin wire dipole antennas using numerical methods.
Visualized the properties and methods (method of moments) using MATLAB.
A two-player LAN game written in python. Inspired by Miniclip's Anagrammatic
Random Python Scripts
Several small python scripts that I develop to automate some of my day to
day tasks. Hosted on github
Development of Database Management System for Essar Power Township, Mahan
Created a database management system (DBMS) for the residents of the company
Protection Schemes of a Thermal Power Plant
Studied and analyzed the various types of electrical faults and protection
schemes implemented in the thermal power plant to prevent any type of damage.
Analysis of the Electrostatic Precipitator and Its Controller
Studied in detail the working of an ESP and analyzed the functioning of its
various components Also examined the ESP Controller used in the plant.