Google releases SpeciesNet, an AI model to identify wildlife

Google has open sourced SpeciesNet, an artificial intelligence model designed to automatically identify animal species from photos captured by camera traps.

The new tool is expected to accelerate wildlife research by dramatically reducing the time it takes to sift through the massive volumes of data generated by these devices.

Camera traps—digital cameras equipped with infrared sensors—are widely used by researchers around the world to monitor wildlife populations. However, the sheer quantity of images they produce can delay data analysis by days or even weeks. SpeciesNet addresses this challenge by classifying images into more than 2,000 labels, ranging from specific animal species and broader taxa such as “mammalian” or “Felidae” to non-animal objects like vehicles.

The model is a key component of Google’s Wildlife Insights platform, an initiative launched about six years ago under the Google Earth Outreach philanthropy program. Wildlife Insights enables researchers to share, identify, and analyze wildlife images online, fostering collaboration to speed up data processing. SpeciesNet was trained on over 65 million publicly available images, supplemented by data from organizations such as the Smithsonian Conservation Biology Institute, the Wildlife Conservation Society, the North Carolina Museum of Natural Sciences, and the Zoological Society of London.

In a blog post published Monday, Google stated, “The SpeciesNet AI model release will enable tool developers, academics, and biodiversity-related startups to scale monitoring of biodiversity in natural areas.” The model is now available on GitHub under an Apache 2.0 license, allowing commercial use with minimal restrictions.

While Google’s release is not the only open-source solution for automating camera trap analysis—Microsoft’s AI for Good Lab has developed PyTorch Wildlife, an AI framework for animal detection and classification—SpeciesNet marks a significant step in democratizing access to advanced AI tools for wildlife research.

By streamlining the analysis of camera trap data, SpeciesNet is poised to support conservation efforts and provide deeper insights into global biodiversity, ultimately aiding researchers in monitoring and protecting ecosystems more effectively.

Copied