Google Unveils SpeciesNet: AI for Wildlife Identification

Google has recently made SpeciesNet, an AI model, open source. This tool is designed to identify various animal species by analyzing photos captured by camera traps, which are essentially digital cameras equipped with infrared sensors used by researchers globally to monitor wildlife populations. The challenge with these camera traps is the sheer volume of data they produce, which can take a considerable amount of time to analyze.

To address this issue, Google introduced Wildlife Insights about six years ago as part of their Google Earth Outreach philanthropy program. This platform allows researchers to upload, share, and analyze wildlife images collaboratively, significantly speeding up the process of data analysis from camera traps.

SpeciesNet, which powers many of the analysis tools on Wildlife Insights, was trained on an impressive dataset of over 65 million images. These images were sourced from public collections as well as from notable 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.

According to Google, SpeciesNet is capable of classifying images into over 2,000 different labels. These labels not only cover specific animal species but also broader taxa like "mammalian" or "Felidae," and even non-animal objects such as "vehicle."

In a blog post released on Monday, Google highlighted that the release of SpeciesNet as an open-source tool will empower tool developers, academics, and biodiversity-focused startups to enhance their monitoring of biodiversity in natural environments. The model is available on GitHub under an Apache 2.0 license, which means it can be used commercially with minimal restrictions.

It's important to mention that Google isn't the only player in this field. Microsoft's AI for Good Lab also offers PyTorch Wildlife, another open-source AI framework that includes pre-trained models specifically fine-tuned for detecting and classifying animals in camera trap images.