Lightweight encoder-decoder network achieving real-time semantic segmentation on urban driving scenes at 38 FPS.
- Computer Vision
- Deep Learning
- PyTorch
A collection of my favorite projects and research in different areas that I have worked on.
Research and coursework from my bachelor’s and master’s programs.
Lightweight encoder-decoder network achieving real-time semantic segmentation on urban driving scenes at 38 FPS.
Fast arbitrary style transfer using adaptive instance normalization, supporting real-time video stylization at 20ms per frame.
Side projects, research explorations, and experiments I have done independently.
More projects coming soon.
Some funny games that I built in my free time.
Interactive ray marching renderer running entirely in a GLSL fragment shader, with procedural terrain, soft shadows, and atmospheric fog.
Simple web application built with HTML, CSS, and JavaScript that simulates a dice roll whenever the page is refreshed. It was originally created as a practice project for a web development bootcamp. h