Varun Thirupathy

Dual Degree Student | AI/ML & Robotics Researcher
Kharagpur, IN.

About

Highly analytical and results-driven Dual Degree student specializing in Electronics and Electrical Communication Engineering, with a robust foundation in Deep Learning, Computer Vision, and Robotics. Proven ability to innovate and deliver high-impact solutions, demonstrated by a publication accepted at IEEE ICDCS 2025 and significant contributions to projects in autonomous driving and federated learning. Seeking to apply advanced research capabilities and technical expertise to develop cutting-edge AI/ML systems and drive technological advancements in dynamic environments.

Work

HiRo Lab (IISc Bangalore)
|

Research Intern

Bangalore, Karnataka, India

Summary

Led the deployment and evaluation of advanced diffusion policies for robotic arm control, significantly improving accuracy and action diversity.

Highlights

Deployed diffusion policies utilizing FiLM-ResNet and temporal U-Net for real-time 1 kHz control of a robotic arm, achieving 92.14% test-time accuracy.

Collected and structured over 130 RGB-D frames and action trajectories into ReplayBuffers, enabling robust training and evaluation of robotic control models.

Improved action diversity by 35% over IBC and BET baselines using multi-modal sampling with Diffusion Policy, enhancing robotic arm adaptability.

Modified the evaluation script from an RTDE-based interface to the ROS2 Framework, resulting in 92% test-accuracy for the Franka arm and streamlining research workflows.

Volunteer

Students Welfare Program (ECE department)
|

Mentor

Kharagpur, West Bengal, India

Summary

Mentored first-year students, providing academic and professional guidance within the ECE department's welfare program.

Highlights

Mentored 4 first-year students from the ECE department, providing academic guidance and fostering professional development within the Students Welfare Program.

Department of Electronics and Electrical Communication Engineering
|

Junior Secretary

Kharagpur, West Bengal, India

Summary

Served as Junior Secretary, contributing to departmental operations and student engagement initiatives.

Highlights

Managed administrative tasks and supported departmental events, enhancing operational efficiency and student engagement within the ECE department.

Autonomous Ground Vehicle Research Group
|

Vision and Controls Team Member

Kharagpur, West Bengal, India

Summary

Contributes to cutting-edge research in autonomous ground vehicles, leading to significant publications and simulations.

Highlights

Collaborated with lead researchers on Vertical Federated Learning, contributing to a publication accepted at IEEE ICDCS.

Participated in CDC 2025, implementing Pure Pursuit and Stanley controllers for autonomous F1Tenth car simulation, demonstrating advanced control system expertise.

Education

Indian Institute of Technology Kharagpur
Kharagpur, West Bengal, India

Dual Degree (BTech and MTech)

Electronics and Electrical Communication Engineering

Grade: 8.13/10 CGPA

Courses

Advanced Calculus

Probability and Statistics

Linear Algebra

Optimization Models

Signals and Systems

Numerical and Complex Analysis

Programming and Data Structures

Digital Signal Processing

Control Systems

Machine Learning specialization (Coursera)

CS-182 Deep Learning (UC Berkeley)

Summer Analytics (IIT Guhwathi)

Publications

Mixup-VFL: Leveraging Unaligned Data for Enhanced Regression in Vertical Federated Learning

Published by

HotDiML @ IEEE ICDCS 2025

Summary

Co-authored a research paper accepted at HotDiML @ IEEE ICDCS 2025, Glasgow, UK, focusing on novel methods for enhanced regression in Vertical Federated Learning.

Skills

Programming Languages

Python, C/C++, Arduino, Verilog, MATLAB, LaTeX.

Libraries & Frameworks

NumPy, Pandas, Matplotlib, Seaborn, Scikit-learn, PyTorch, TensorFlow, OpenCV, ROS1/ROS2.

Tools & Software

Google Colab, Kaggle, Docker, Git, PyCharm, VSCode, Proteus, Xilinx, Vivado, LTSpice.

Technical Domains

Deep Learning, Natural Language Processing (NLP), Convolutional Neural Networks (CNN), Computer Vision, Federated Learning, Generative AI (VAE, GAN, Diffusion Models).

Projects

Stanley Controller - F1Tenth Pipeline

Summary

Developed a geometric path tracking algorithm for small-scale autonomous vehicles, specifically for F1Tenth cars.

Mixup-VFL: Leveraging Unaligned Data in Vertical Federated Learning

Summary

Developed innovative Mixup methods to prevent data wastage in Vertical Federated Learning settings, enhancing multi-party learning.

Neural Network Implementation from Scratch

Summary

Implemented Deep Learning models from scratch to reinforce mathematical foundations and gain a deeper understanding of neural network mechanics.