An AI model called SpeciesNet is released by Google to identify wildlife

An AI model called SpeciesNet is released by Google to identify wildlife

Google releases SpeciesNet, an AI model that analyzes images from camera traps to identify animal species.

 
Camera traps, which are digital cameras attached to infrared sensors, are used by researchers worldwide to examine wildlife populations. However, even though these traps might yield insightful information, they also produce enormous amounts of data that require days or weeks to sort through. About six years ago, as part of its Google Earth Outreach philanthropic program, Google started Wildlife Insights as a way to try to help.
 
Researchers can collaborate to expedite camera trap data processing by sharing, identifying, and analyzing wildlife photos online using Wildlife Insights.
 
SpeciesNet, which Google claims was trained on more than 65 million publicly accessible images as well as images from institutions like the Smithsonian Conservation Biology Institute, the Wildlife Conservation Society, the North Carolina Museum of Natural Sciences, and the Zoological Society of London, powers a large number of Wildlife Insights' analysis tools.
 
According to Google, SpeciesNet can categorize photos into over 2,000 labels, including non-animal things (like "objects"), taxonomies like "mammalian" or "Felidae," and animal species.
 
In a blog post, Google stated that the SpeciesNet AI model release would allow researchers, tool developers, and companies involved in biodiversity to scale biodiversity monitoring in natural areas.