This project was done during my work with Kanastruk
The goal was to provide a tool that would estimate the total volume of objects based on a photo.
A first version is available on the ramasse.ca website, using LLMs to recognize, list and estimate the volume of objects.
Another prototype is in the works, using both LLMs and Computer Vision models to allow the user to manually select objects.
demo of the computer vision prototype coming soon...
Designed by HTML Codex