Introductory

Computer Vision for Lunar Rover Navigation

In collaboration with the Robotics Club, we’re developing a computer vision algorithm to enable lunar surface rovers to detect ice, avoid obstacles, and map terrain accurately using cameras, LiDAR, and machine learning techniques like SLAM and CNN.

Advanced

Game Diffusion Model

We’re building a game engine that uses a near real-time diffusion model to generate game elements based on user-specified image segmentation masks, allowing for dynamic and interactive map customization.

Music Transformer

We’re developing an ML-driven chord generator that helps musicians create harmonious chord progressions by analyzing previous MIDI input and offering optimal chord suggestions based on their desired musical feel.

LLM Context Highlighting

We’re building a large language model to identify and highlight semantically related tokens in a text editor, enhancing the context and depth of writing by connecting conceptually linked words and phrases.