About Me
I am currently pursuing a Master’s degree in Robotics at Carnegie Mellon University. Previously I interned at Raven Industries, where I focused on leveraging Computer Vision and AI to enhance Autonomous Farming. Additionally, I worked as a Research Engineer at The Hi-tech Robotic Systemz, where my efforts were dedicated to advancing AI and Robotics technologies for on-road logistics and intralogistics solutions.
My interests include Computer Vision, Natural Language Processing, Deep Learning, Machine Learning, EdgeAI, Parallel Programming, Robot Perception, Sensor Fusion, and more.
I am very keen towards application-driven research, and fusing fundamental research into solving real-world problems has been my biggest motivation.
Download my resumé.
- Computer Vision
- Deep Learning
- Machine Learning
- Artificial Intelligence
- Natural Language Processing
- Mobile Robotics
Master of Science in Robotic Systems Development (MRSD), May 2024
Carnegie Mellon University, School of Computer Science
Bridge Program in Computer Science
New York University
B.Tech in Mechatronics
SRM University
Experience
ADAS and Autonomous Driving:
- Integrated a RADAR to Novus Aware device to add a Forward Collision Warning (FCW) feature. Created a Qt-based GUI tool to visualize objects from Radar, Improved the collision warning to be robust using Bayesian filters.
- Developed a novel HM-LSTM model to predict early drowsiness using blink features, optimized using TFlite, and deployed it using ArmNN on low-cost ARM hardware.
- Improved the speed of object detection models by 2.5X using quantization, model tweaking, and TensorRT without any accuracy loss.
- Developed a MobileNetV3 backboned multi-branch network to detect vehicles and segment lanes simultaneously.
- Developed and deployed a Real-Time Traffic Light Detection model with 59% mAP for our autonomous driving shuttle.
- Prototyped and developed 6 camera Surround View System (SVS) for commercial vehicles.
- Hands on experience training and deploying various CNN models (MobileNet, ResNet, SSD, YOLO) and LSTM based models.
IoTization with AWS IoT:
- Built a real-time web streaming feature (WebRTC) for Novus Aware device to stream driving videos to clients.
- Improved the driving data collection pipeline using AWS IoT MQTT to be more cost-effective.
- Implemented and Automated IoT device registration for our fleet, reducing significant manual effort.
Mobile Robotics:
- Built a timed mission feature so that the robot could perform a specified task at a given time and integrated it with the UI.
- Improved robot localization and loop closure significantly by fusing an IMU using EKF to existing RTAB-Map localization.
- Built and integrated AR Tag detection algorithm to enable fiducial marker-based navigation for mobile robots.
- Designed various UI features for the Robot fleet UI to enhance the user experience of the clients using Ros2djs.
ADAS Team:
- Developed an object detection model with 54% mAP to detect vehicles and pedestrians for Indian Scenarios.
- Deployed the Object Detection model using TensorRT on a Jetson TX1 with 23 FPS.
- Dealt with class imbalance using Focal Loss and significantly improved object detection accuracy.
- Implemented a verification algorithm to validate the objects from radar and camera with ground truth efficiently.
Projects
Certifications
