Resume

Neocis Inc, Software Engineer Intern (May 2022 - August 2022)

• Designed algorithm to detect dental implants in CBCT Scan point cloud data and compute error metrics post-robot-assisted dental operation. Improved error analysis accuracy and consistency over existing methods by 27%.

• Researched singularity and joint limit avoidance algorithms for a 6 DOF robot arm.

Deepedge AI, Machine Learning Engineer (January 2021 - July 2022)

• Collaborated with Luum Robotics on Lash robot's AI and Vision system for extending eyelashes in beauty salons.

• Developed and deployed a Siamese network-based model for one-shot image recognition of identical eyelashes in stereo image pairs using TensorFlow in C++, achieving 98% accuracy in stereo matching and reducing inference time by over 200ms.

• Developed auto-annotation pipeline, enhancing the data labeling process and reducing the turnaround time by nearly 50%.

ICICI Bank, Management Trainee (October 2020 - December 2020)

• Managed the marketing data from the various internet banking touchpoints, and developed meaningful insights to target customers for all the products in the ICICI Bank portfolio to boost sales.

Mercedes-Benz Research and Development India, Student Trainee (Januray 2020 - July 2020)

• Initiated experimental project with Noise, Vibration, and Harshness Team involving analysis of noise signals and classification using fingerprinting and Neural Networks.

• Created an ESResNet based diagnostic tool using Pytorch to classify various vehicle noises, predict their source, and eventually find a serviceable solution for the problem.

• Minimized diagnosis and resolution time for service technician, to consequently improve customer experience by reducing overall servicing time.

UAV Mapping with Semantic and Traversability Metrics for Forest Fire Mitigation (January 2021 - May 2022)

• Formulated fuzzy logic-based algorithm to determine ground traversability for UGV using point cloud data from UAV as part of SafeForest project.

• Together with research group implemented SLAM system to understand 3D structure of forest environment, a semantics system to identify fuel and other features in environment, and a system that predicts which region a UGV can navigate.

  • Paper in OpenReview and oral presentation in ICRA 2022 Workshop IFRRIA.

Context-based Video Colorization (March 2022 - July 2022)

• Building a novel, fully automatic neural network architecture with context-based attention to colorize grayscale videos by augmenting colorful image colorization for temporal coherency.

Automatic Eye-in-Hand Calibration using EKF (March 2022 - May 2022)

• Conceptualized employing Extended Kalman Filter to automatically calibrate an industrial robotic arm with a camera on end-effector to make hand-in-eye calibration fast and effortless.

Aurecle - Augmented Reality Clearance Estimation (November 2021 - December 2021)

• Developed a computer vision-based warning system for logistic vehicles to estimate clearance height under bridges and overpasses to avoid over height vehicle crashes.

KickFlip - The Cooking Robot (August 2021 - December 2021)

• Built a kitchen appliance cooking robot prototype with an integrated cooktop, 2-DOF actuated spatula, and sensing probe to automatically cook burger patties to the desired cooking level.

• Integrated various sensors to get density and temperature and to observe texture and colour to monitor cooking status and actuate relevant mechanisms.

Control and Planning for Vehicle and Quadcopter (September 2021 - December 2021)

• Applied PID and LQR controllers, EKF for localization and A* planning for lane change using Python on vehicle and quadcopter in Webots simulator.