top of page
Search

AI Computer Vision with MoonDream

Code and Workbook on Github


Description

Computer Vision no longer requires a PhD and a high powered computer. With some Python code and a potato you can bring that power of Computer Vision to your projects.


Whether you want to categorize pictures, build a surveillance system that responds in real time, or build a robot that can "see" the MoonDream Vision Learning Model can quickly and easily be integrated into your projects.


It's built to be lightweight so that it can run on edge systems, AND it can provide X and Y coordinates for objects which most models can't. By having the coordinates of objects you can do some pretty cool stuff.


We'll explain how MoonDream works and then integrate it into a number of projects to show how it can be used.


This class will go over:

  • What is MoonDream

  • How to setup and run MoonDream

  • Understanding Architecture when designing systems that use Vision.

  • Query - Ask queries about an image

  • Detect - Detect if objects are in an image

  • Point - Find the X/Y coordinates of objects in images

  • Caption - Create captions for images.


We will explain how these services work. Demonstrate how to use these services in live code, and you will have time to create simple labs to get the feel for how easy AI can be.


bottom of page