At Larsen & Turbo, I collaborated with a team on a construction-site safety-gear detection system. We implemented a two-stage Faster R-CNN model to detect workers’ helmets and jackets, and analyzed datasets to achieve 99% classification accuracy.
As an Academic Intern at National University of Singapore, I collaborated on a liver tumor segmentation system using clinical and CT scan data. I performed exploratory data analysis and feature engineering, implemented ensemble learning that achieved 99% classification accuracy, and deployed the project using AWS services.
In my projects, I trained a pneumonia detection model using CNN transfer learning and built a German traffic sign classifier that achieved 99% test accuracy. I also used historical Formula1 data to build a linear regression model for predicting qualifying times and the winner.

