Google has made available to the public the SpeciesNet artificial intelligence model designed to identify animal species by analyzing photos from camera traps.
Researchers around the world use camera traps – digital cameras connected to infrared sensors – to study wildlife populations. But while these traps can provide valuable information, they generate massive amounts of data that can take days to weeks to sift through.
In an effort to help, about six years ago, Google launched Wildlife Insights, an initiative of the Google Earth Outreach charity program. Wildlife Insights provides a platform where researchers can share, identify, and analyze wildlife imagery online, collaborating to speed up the analysis of data collected through camera traps.
Many of Wildlife Insights’ analysis tools are powered by SpeciesNet, which Google claims has been trained on more than 65 million publicly available images and images 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.
Google says that SpeciesNet can classify images by one of more than 2,000 labels covering animal species, taxa such as “mammal” or “feline”, and non-animal objects (such as “vehicle”).
“The release of the SpeciesNet AI model will enable tool developers, scientists, and biodiversity-related startups to scale biodiversity monitoring in natural areas,” Google wrote in a blog post published on Monday.
SpeciesNet is available on GitHub under the Apache 2.0 license, which means it can be used for commercial purposes with virtually no restrictions.
It is worth noting that Google is not the only open source tool for automating the analysis of trap camera images. Microsoft’s AI for Good lab supports PyTorch Wildlife, an artificial intelligence framework that offers pre-trained models configured to detect and classify animals.
