Researchers around the world use camera traps — digital cameras connected to infrared sensors — to study wildlife populations. While these traps provide valuable insights, the sheer volume of data they generate can take days or even weeks to sort through.
To address this challenge, Google launched the Wildlife Insights project about six years ago as part of its Google Earth Outreach charitable initiative. Wildlife Insights offers an online platform that allows researchers to share, identify, and analyze wildlife images, accelerating the processing of camera trap data through collaboration.
Many of the analytical tools within Wildlife Insights are powered by SpeciesNet. According to Google, SpeciesNet has been trained on over 65 million publicly available images, as well as images from organizations such as the Smithsonian Conservation Biology Institute, Wildlife Conservation Society, North Carolina Museum of Natural Sciences, and the Zoological Society of London.
SpeciesNet is capable of classifying images with more than 2,000 labels, covering animal species, taxonomic groups like "mammals" or "felines," and non-animal objects such as "vehicles."
In a recent blog post, Google announced the release of the SpeciesNet AI model, enabling tool developers, academics, and biodiversity-focused startups to scale up biodiversity monitoring in natural areas.
SpeciesNet has been open-sourced on GitHub under the Apache 2.0 license, allowing it to be used for commercial purposes with minimal restrictions.
It's worth noting that Google isn't the only organization offering open-source tools for automated analysis of camera trap images. Microsoft’s AI for Good Lab maintains the PyTorch Wildlife framework, which provides pre-trained models fine-tuned for animal detection and classification tasks.
