My Projects

Explore my work in robotics and machine learning, from interactive learning frameworks to autonomous navigation systems.

PHARMore, Data Mining for Drug Discovery

Designed a research support tool for drug repurposing that combines automated information extraction, intelligent recommendation, and conversational assistance. The system streamlines literature analysis and hypothesis generation using NLP, biomedical databases, and an AI-powered chatbot interface. The system streamlines literature analysis and hypothesis generation using NLP, biomedical databases, and an AI-powered chatbot interface.

PythonStreamlitAWSGeminiPyTorch

Feb 2025 - Feb 2025

HINTeract, Interactive Robot Learning Framework

Trained Behavioral Cloning policies from 260+ demonstrations for subtask-level table assembly in Robosuite. Built a hierarchical imitation learning framework with hint-guided feedback, optimizing sample efficiency and task rollout stability.

PythonRobosuitePyTorch

Jan 2025 - Apr 2025

SpooderMan, Autonomous Robot Follower

Developed an autonomous robot follower on TurtleBot3 by employing SIFT-based feature extraction and image segmentation using OpenCV and custom ROS2 packages Integrated a SLAM framework with a dynamic window-based controller for precise global localization and coordinate-based navigation, enabling efficient maze traversal through KNN-based directional sign classification

ROS2PythonOpenCV

Aug 2024 - Dec 2024

Oogway, Autonomous Rover and Robot Companion

Developed a SLAM system on TurtleBot 2 using SURF-based feature detection, enabling reliable recognition and tracking of over five distinct objects in a maze environment with ROS, gmapping, and OpenCV. Integrated object-following and emotion-display modules that resulted in an 80% user satisfaction rating, improving the system's responsiveness and user engagement.

ROSC++OpenCV

Jan 2024 - Apr 2024

Kir-B, Autonomous Maze-Navigating Rover

Designed an Arduino-based autonomous maze-navigating rover, achieving a top 3 ranking for autonomy in competition. Developed and integrated modules for obstacle avoidance, localization, object detection, and communication, enabling efficient sensor data transfer from Arduino Mega via Bluetooth and high-level command processing in Python. Improved localization accuracy by 70% through advanced sensor data preprocessing and the implementation of a particle filter algorithm.

PythonArduino

Sep 2023 - Dec 2023

Driver Drowsiness Detector, Real-time Drowsiness Detection

Designed and trained a convolutional neural network (CNN) in PyTorch for high-accuracy driver drowsiness detection (98.9%), with real-time facial feature extraction from video frames using OpenCV. Implemented a bounding box system for precise feature detection, achieving reliable monitoring in live video scenarios.

PyTorchOpenCV

Jan 2021 - Apr 2021

Interested in learning more about my background and experience?

Learn More About Me