Selected Engineering Projects

Swarath Autonomous Car

Guide: Dr. Saket Anand, Dr. Sanjit Kaul, Dr. Rahul Purandre
Team Size: 3
Team Member(s): Shivam Thukral, Gurshabad Grover
Technologies/Concepts: ROS, C++, Design Patterns

As a member of IIIT-Delhi’s core team participating in Mahindra’s National Driver Less Challenge, I designed Swarath’s test suite and perception module in ROS/C++. The test suite allowed for creation and automatic evalutation of testing sets for specific development tasks. The perception module followed the strategy design pattern and externalized control to the main dashboard, enabling switching between algorithms and sensors for core perception tasks such as path planning, object detection, 3D mapping and lane detection.

May'15-May'16

Fettle Human Computer Interaction

Guide: Dr. Ponnurangam Kumaraguru
Team Size: 6
Team Member(s): Saumya Jain, Aditi Mithal, Manan Gakhar, Danish Goel, Sumit Keswani
Course: Designing Human Centered Systems
Technologies/Concepts: Android, User Interviews, Affinity Diagrams, High/Low Fidility Prototyping, Design Principles

Fettle is a highly intuitive food recommendation cum fitness tracking android application. Besides just allowing a user to order food from nearby restaurants or track their physical activities, the application closed the loop by integrating both the experiences and enabled a user to stay in control of his/her health. Based on a person’s physique/targeted physique on the food ordering menu, the application sorts the food items based on a user calorie requirements for the day. The system also tracked the fitness activity of the user and allowed for this food ranking to get relaxed if the person had been working out well. Alongside all the functionality, the application was highly intuitive and minimal in design, completely abstracting the user from the underlying complexity and not overloading him/her with these functions.

Jan'16-May'16

Meri Awaaz Web Dashboard

Team Size: 5
Team Member(s): Manan Gakhar, Danish Goel, Shashank Gautam, Anchita Goel, Mansi Panwar, Aditya Kumar Nayak
Technologies/Concepts: Javascript, JQuery, Jekyll, Foundation, Parse Cloud Computing, REST API Development

Meri Awaaz is India’s first platform that connects citizens with their netas and government officials for Open, Accessible and Accountable Governance. I developed the web dashboard which was used by candidates/netas in two constituencies of Delhi to interact with the respective citizens and address their issues. The dashboard allowed elected officials to view citizen issues, delegate work to their team, run campaigns, post updates about the progress made and compare their efficiency with other politicians within the constituency.

Aug'14-Dec'15

Kinect based 2-D Mesh Animation Computer Graphics

Accepted Project Demo At IIIT-Delhi's Research Showcase 2016
Guide: Dr. Ojaswa Sharma
Team Size: 1
Course: Computer Graphics
Technologies/Concepts: QT, OpenGL, Kinect, Visual Studio

In this project I focus on animating a 2D mesh character and driving its actions through kinect. The system implemented in Visual Studio interact with our kinect device and fetch the skeleton data captured in the form of coordinates. We then create our 2D mesh in QT and use these coordinates to mimic the skeleton. The project follows a bone based skeleton model to construct our 2D mesh character. The project can seamlessly by ported to work for 3D characters as well.

Aug'15-Dec'15

coDrive Crowd Souced Memory Sharing

Guide: Dr. Rahul Purandre
Team Size: 4
Team Member(s): Manan Gakhar, Danish Goel, Chaitanya Kumar
Course: Practice of Programming
Technologies/Concepts: Android, Parse, Javascript, Lint, SonarQube

Often offline storage runs out on mobile devices, and the user then needs to store data either on cloud, or on external media. We propose a solution, wherein the user needn’t be concerned about the security of their data in the cloud, by simply storing it on a dedicated space on one of his trusted peers, at the luxury of a network. We have implemented an application that allows multiple users to share some space on each of their drives (offline storages). This is achieved by initially setting up a fixed amount of space on each user’s local drive, required for the purpose of the application. Once a user issues a request to his/her list of “friends”, it can be accepted for storing the user’s data.

Jan'15-May'15

NEWSSS - Social Search

Guide: Dr. Sameep Mehta
Team Size: 2
Team Member(s): Saumya Jain
Course: Information Retrieval
Technologies/Concepts: Page Rank, NLP, Social Media Analytics

Social search is a behaviour of retrieving and searching on a social searching engine that mainly searches user-generated content such as news, videos and images related search queries on social media. In this project we create our own Social Search Engine implementing Adapted PageRank Algorithm based on the paper - Information retrieval in folksonomies: Search and ranking, A Hotho, R Jäschke, G Stumme

Jan'16-May'16

Color based Offside Detection Blob Detection

Guide: Dr. Mayank Vatsa
Team Size: 1
Course: Image Analysis
Technologies/Concepts: C#, EmguCV, Visual Studio 13, Image Processing

As a part of my minor project I built a software in C# using emguCV Library to detect players in a football field using Color based object Segmentation. I further wrote an algorithm to detect offside in the game by combining data coming from cameras located at various angles across the field.

Aug'14-Dec'14

reWise In-Class Quizzing Application

Guide: Dr. Vinayak Nayak
Team Size: 5
Team Member(s): Saumya Jain, Shuktika Jain, Mrinal Kachara, Manan Gakhar
Course: Mobile Computing
Technologies/Concepts: Android, Parse, Javascript

An android application to hold quick in-class quizzes. In our project we implement an intuitive UI, a persistent question bank, statistical analysis of quizzes as well as individual performances and gamification using badges.

Aug'14-Dec'14

He-or-She Speech based Gender Detection

Guide: Dr. Richa Singh
Team Size: 2
Team Member(s): Anchita Goel
Course: Pattern Recognition (Minor)
Technologies/Concepts: Matlab, Speech Signal Processing, Supervised Learning

In this project we implement a system to detect the gender of the speaker based on various features like MFCC, pitch, short-time energy, energy entropy, zero-crossing rate and spectral centroid. We also implemented an android application for the same and report a detailed evaluation of the results on a public dataset.

Jan'15-May'15

Image Recolorization Chrominance Blending

Guide: Dr. Mayank Vatsa
Team Size: 2
Team Member(s): Manan Gakhar
Course: Image Analysis (Major Project)
Technologies/Concepts: Arduino, IR & Ultrasonic Sensors

As a part of our Major Project our team, implemented and analysed the performance of Chrominance Blending Technique proposed by Liron Yatziv and Guillermo Sapiro in their paper on the same comparing it’s accuracy and computation speed with other common approaches.

Apr'14-Aug'14

Scene Recognition using Bag-of-Words Machine Learning

Guide: Dr. Richa Singh
Team Size: 2
Team Member(s): Anchita Goel
Course: Machine Learning
Technologies/Concepts: Matlab

Scene recognition is an important real world problem which finds application in the fields of robotics(autonomous systems), surveillance(wearable camera footage, evidence photographs, etc), and personal assistance devices(Google Glass). There has been a steady progress in the field and this mini-project explains the developments made leading to the current state-of-the-art approach, documenting the results obatined from our implementation with standard benchmarks and classifiers.

Aug'15-Dec'15

Pedestrian Detection based on HoG

Guide: Dr. Saket Anand
Team Size: 2
Team Member(s): Prateekshit Pandey
Course: Computer Vision
Technologies/Concepts: Matlab, Computer Vision Toolbox

Human Detection in Images is a contemporary Computer Vision problem. This subset area of object detection has seen many attempts made towards improving the accuracy and in this project we implement from scratch one based on Histogram of Oriented Gradients which continues to be one of the most used methods in terms of both detection rate and error rate.

Jan'15-May'15

Travel Booking System SAAS

Guide: Mr. Manish Shrotriya
Team Size: 8
Team Member(s): Saumya Jain, Shuktika Jain, Mrinal Kachhara, Manan Gakhar, Mansi Verma, Akanksha, Danish Goel
Course: Software Engineering
Technologies/Concepts: Ruby on Rails, Javascript, AGILE Paradigm

As a part of the Software Engineering course we made a SAAS web application using ruby on rails framework following the agile development. The Web Application consisted of a system to allow a Travel Agent to effectively run his business by adding and using a set of applets that the service provides.

Aug'14-Dec'14