top of page

Research

Live Environment Object Recognition and Model Correspondence

Suraj Karthikeyan, Sam Hus, Jay Ho, David Khankin
Fall 2023

This project was developed for Michigan State University's Computer Vision graduate course. The end goal of the project was the development of a effective prototype using pretrained models in real world locations.

Using a model uploaded by the user, they can select an object from their live camera feed and compare this against the model. This program will highlight any instances the object in the corresponding model.

This project makes use of BASNet, SuperPoint and LightGlue.

bottom of page