Critterpedia: an AI-powered app to identify insect and snake species

July 15th, 2020

Critterpedia is an AI-powered app to identify insect and snake utilising photos. The platform uses an algorithmic solution to identify the species of spider or snake submitted by users.

By utilising a photo taken with any smart device, a trained algorithm classifies it and provides information on the family, genus or species of the animal.

In order to do so, the platform required hundreds of thousands of images of snakes and spiders to be fed into the system to accurately establish Australia’s multitude of species, a sizeable and manual task uniquely suited to an artificially intelligent (AI) solution.

In partnership with CSIRO’s Data61 Imaging and Computer Vision Group, the Critterpedia app will have a machine learning engine for automated species identification embedded, allowing it to rapidly and seamlessly identify, sort, and certify large volumes of complex information.

Introduced to our team via CSIRO’s Kick–Start team, which helps start-ups find research and development (R&D) capabilities and provides dollar-matched funding for projects. The adoption of AI will ensure Critterpedia’s search engine and recognition system is maximised for accuracy, a critical consideration when it comes to pinpointing the finer differences between species which may or may not be deadly.

“The visual differences between two species can sometimes be quite subtle, and so a great deal of training data is needed to adequately identify critters,” explains project lead and Data61 researcher, Dr Matt Adcock.

“We’ve started off with an enormous amount of images sourced from zoological experts collaborating with Critterpedia, and have developed a suite of tools to help semi-automatically label these images, verify the information, and cross check with other data sources.”

“The AI platform we are developing for the Critterpedia system considers not only these images, but also additional information, such as GPS location.”

Users can contribute to the Critterpedia system by submitting photos. The more differences and variables provided will ensure the robustness of the identification and information system.

Users can sign up to become a Phase 1 tester, which lets you download a beta version of the platform and submit wildlife photos to keep training the algorithm.

As a wildlife safety, awareness and education hub, Critterpedia aims to provide better education and awareness for all Australians across a number of sectors and could ultimately save human and animal lives.

The Critterpedia team have been fortunate enough to align with CSIRO/Data61, Ignite Alliance, SingularityU, Josephmark, Slipsteam Commercialisation, Advance Queensland, and Startup Onramp, who have guided them through their journey so far. Also, with a special thanks to their passionate snake and spider advisors, including Robert Whyte, Scott Johnson, Narelle Murphy, Ben Shoard, Kane Durrant and Scott and Tie Eipper, their business mentor, Lewe Atkinson, and business confidant, Megan Avard (SurePact). Not to mention the 30 incredible snake and spider experts, including Iain Macaulay, Paul Irvine, Adam Brice, Mick Fullerton and John and Tina Mostyn, who have shared their images to aid in the training of the machine learning AI algorithm.

Read more on Algorithm.

Media coverage

Contact us to learn more.

Subscribe to our News via Email

Enter your email address to subscribe and receive notifications of new posts by email.