WALDO

Hey HN! I just released a suite of AI models for deployment on UAV and other "overhead" devices to provide some understanding of the world below. The objective is to empower all sorts of open-source use cases around search and rescue, wildfire prevention, ground risk mitigation for flight over populated areas etc... The neural networks are trained for a bunch of different devices from big GPUs to tiny edge AI cameras like the Luxonis OAK, with some optimised ones for Nvidia TensorRT and other cool bits and pieces too. The main release package also includes some boilerplate code for running on your own drone images and videos, so check it out whether you have a DJI hobby drone or a massive Raytheon platform!

First release of WALDO: Whereabouts Ascertainment for Low-lying Detectable Objects (get it?) This is a super-light, fast & basic single-shot-detector trained for UAV to detect and avoid objects on the ground during flight based on a YoloV7 model and my own synthetic drone flight datasets! This first generation supports a single output class (people) and is meant to be used as a completely free security feature for anyone building UAV whether for fun or for commercial purposes! This is alpha 0.1 so very early days but just like OpenLander the code is 100% FOSS and usable for commercial deployments from day 1. There's a lot of progress still to be made on the detection but this first generation works surprisingly well... If you're more interested in segmentation check out my other repo for UAV AI called "OpenLander" :)