Conservation AI aims to harness machine learning for various conservation projects. At present we focus on detecting and classifying animals, humans, and man-made objects indicative of poaching (e.g. cars, fires). We focus work with images from visual spectrum and thermal infrared cameras that are used on drones or in camera traps. The aim is to provide a user-friendly workflow that can allow for near-real time detection/classification and non-real time detection/classification.
We have ongoing test projects with Knowsley Safari in the UK, the Endangered Wildlife Trust in South Africa, and the Greater Mahale Ecosystem Research and Conservation team.
When images, video or audio files are uploaded to the Conservation AI site they are automatically classified by our trained models to return the species and the associated probability value. Images and audio files can be sent automatically from a variety of different sources including camera traps, drones and existing media. See Getting Started for more details.