Home

About Me

Hey! My name is Aniket Gupta. I am currently pursuing a Master of Science in Robotics from Northeastern University. My concentration in the program is Computer Science.

Prior to joining Northeastern, I completed my undergraduate studies from Delhi Technological University majoring in Electrical Engineering. There I was leading the Team Unmanned Aerial Systems (UAS-DTU) under the supervision of Prof N S Raghava. 

I have broad research interests in computer vision, deep learning, robotics and reinforcement learning. My long term research goal is to develop effective methods to develop visual intelligence in robotic systems with spatial awareness.

Download my Resume here.

Experience

Teaching Assistant, Computer Vision (CS 5330), Northeastern University
  • Responsibilities include helping the students in learning the concepts taught in the class, grading the assignments, exams and projects.
Graduate Research Assistant, Field Robotics Lab, Northeastern University
  • Estimated IMU bias for localization using Dead Reckoning in case of camera failure, achieving a buffer time of above 30 seconds.
  • Explored existing Visual odometry libraries including Kimera and VINSMono to integrate the IMU bias estimation algorithm.
Project Manager, Unmanned Aerial Systems team, DTU
  • Coordinated and managed a multidisciplinary team of 30+ members for multiple applied projects.
  • Developed and Demonstrated long-range (50 km) swarming capability to the Indian Air Force on 25 UAVs and won a project funding of INR 35 crores ($4.7 million).
  • Implemented LeGO-LOAM SLAM on a UGV with factor graphs for improved loop closure.
  • Developed a 3D monocular visual odometry system using deep learning for online depth estimation.
  • Signed an MoU with Adani Defence and Aerospace with a funding of INR 7.5 crores ($1 million).
  • Mentored the students in learning core robotics and deep learning concepts and ensured high research output.
Software Development Intern, Flaire Unmanned Systems
  • Trained 2 scalable CNN models in TensorFlow for aerial reconnaissance at Disaster Struck Areas.
  • Integrated the model in main software pipeline for ‘”Follow Me” mode and unit-tested the code base.
Research Intern, Cyber-Physical Systems Lab, IIITD
  • Developed an AI agent to solve “Travelling Salesman” problem by utilizing traditional Value iteration methods.
  • Designed a Deep Q-Learning-based controller and solved cooperative path planning problem for multiple agents.
  • Assembled and tested complete architecture on a system of 3 agents leveraging Multi-master ROS.
Avionics Technician, Unmanned Aerial Systems team, DTU
  • Developed a Real-time collision avoidance system using a custom-designed stereo-vision system.
  • Added static obstacle avoidance capabilities to a non-holonomic UAV using Artificial potential field algorithm and tested it using the Software in the loop simulation using JSBsim platform.
  • Implemented visual-aided precision landing on a multi-rotor system on a moving platform.
  • Designed and tested the propulsion system of multiple fixed-wing and multi-copter systems, optimized for different criteria like high payload capacity, high endurance or both.

Projects

Deep Learning Projects

Dense Completion from sparse depth maps

[Code] [PDF]

Monocular Depth Estimation

[Code]

Human Detection and Object Tracking

[Code]

Multi-Agent control using Deep RL

[Code] [PDF]

Multi-Agent Control

Swarm flight test – 20 Long range UAVs

[PDF]

UAV Swarm Simulations

[Code] [PDF]

Optimized Search using Multiple Agents

[PDF]

Formation control on Fixed Wing UAV systems

[Code]

Miscellaneous

Bipedal Robot learning to walk using Deep RL

[Code]

Static Obstacle Avoidance for Holonomic bodies

[Code]

Meeting organizer App

[Code]

SLAM on quad-rotor system

[PDF]

Systems Developed

Publications

  1. Aniket Gupta, Parth Mahajan “Depth And Skeleton Based View-invariant Human Action Recognition” IEEE GCAT 2021
  2. Aniket Gupta, Raghava Nallanthigal “Decentralized Multi-Agent Formation Control via Deep Reinforcement Learning“. IEEE Agents and Artificial Intelligence 2021
  3. Aniket Gupta, Parth Mahajan, Aman Virmani, Raghava Nallanthigal “A Particle Swarm Optimization-based Cooperation Method for Multiple-Target Search by Swarm UAVs in Unknown Environments“. Accepted to IEEE International Conference on Automation, Robotics and Application, 2021
  4. Anuj Agrawal, Aniket Gupta, Joyraj Bhowmick, Anurag Singh and Raghava Nallanthighal “A Novel Controller of Multi-Agent System Navigation and Obstacle Avoidance“. Procedia Computer Science, at 7th Symposium on Control, Automation, Industrial Informatics, and Smart Grid (ICAIS’19) Elsevier, 2019.

Awards and Honours

  • Winner – Communication Architecture, Indian Air Force Mehar Baba prize competition.
  • Speaker, “Engineering Robust Aerial Swarms”, Asia-Oceania Systems Engineering Conference, 2019, Bangalore
  • 1st Position, Formation Flying at Drone Olympics 2019.
  • 2nd position, Technical Journal Paper (AUVSI SUAS)
  • 1st position overall and Technical Presentation (SAE India Southern Section AeroDesign Challenge (Regular Class), 2018)
  • 1st position overall and Technical Design Report (SAE India Southern Section AeroDesign Challenge (Micro Class), 2018)
  • National Finalists DRDO Robotics and Unmanned Systems Exposition
  • Awarded Nvidia AI drone development kit by Lockheed Martin.
  • Award of Merit Scholarship, 2018
  • Runner-up, TATA Makerthon challenge, 2018
Get in Touch

Copyright Aniket Gupta – All rights reserved

Design a site like this with WordPress.com
Get started