Now, a new app called Critterpedia aims to act as if a “Shazam” for spiders (and snakes, too), was involved, using a trained Artificial Intelligence algorithm to match its snapshots against a library of suspected arachnids and reptiles.
Since, most likely, if you are in Australia, you wonder moment yes, moment also, if it is a poisonous animal, a team of experts has designed a platform that can quickly identify the family, genus or species of an animal from a photo. The new technology works by scanning a file with hundreds and thousands of photos, accumulating the identification skills of a zoologist with a lifetime of fieldwork experience.
The project is carried out in collaboration with CSIRO’s Data61 with the aim of refining the technology so that Critterpedia provides reliable results about the animal we are dealing with.
” The visual differences between two species can sometimes be quite subtle, so it takes a lot of training data to properly identify the critters,” explained project leader and Data61 researcher Matt Adcock. “We started with a massive amount of images obtained from zoological experts collaborating with Critterpedia, and we have developed a set of tools to help semi-automatically tag these images, verify the information, and cross-check with other data sources .”
Do you want to be a phase 1 tester? Enter here and you can download a beta version of the application.
