Out-of-domain Open Vocabulary Semantic Segmentation
We present Seg-TTO, a novel framework for zero-shot, open-vocabulary semantic segmentation (OVSS), designed to excel in specialized domain tasks. 
Portfolio
Generative AI for Video Analysis
We designed a pipeline to streamline video analysis using LLMs and GenAI. This is associated with my undergraduate internship at SUTD, Singapore. 
Robot Localization using a Single 360-Degree Camera
I developed a computer vision model to calculate the distance and the angle to a robot. This is associated with my undergraduate internship at SUTD, Singapore. 
Autonomous Task Solving Dual Robot System
We Built a dual robot system capable of navigating a maze arena, and a dotted line section, avoiding obstacles, identifying surface colours, and communicating with each other. GitHub 
Opthalmic Biomarker Detection
Research and implementation of a state-of-the-art contrastive learning model and a supervised model to accurately classify the presence of six biomarkers in an OCT scan. This work resulted in the 14th place in the IEEE Video and Image Processing Cup 2023.