One of the first big apps to make us all realize that smartphones were both cool and useful was Shazam. Does anybody else remember the joy of holding up your phone to the song you couldn't get out of your head but also didn't know the name of and finally, finally learning what it was called? So satisfying, right?
Well, researchers in, where else, Australia, have developed a Shazam for creepy crawlies and you never know, it just might save some lives.
The Land Down Under is home to more than 2,000 different kinds of spiders and 170 different snakes. While the vast majority of them pose no real threat to humans, nobody wants to mistake the few that are capable of killing humans for one that's perfectly safe.
That's exactly the idea behind Critterpedia, the app that aims to sort the safe from the deadly, as developed by Australia's National Science Agency (CSIRO).
You see a snake or a spider, you take a pic of said critter, and you upload that pic to the app.
The app uses an AI-powered machine learning algorithm to analyze the image, identify the critter, and provide the user with any relevant info about it, like if it's horrifically venomous or not.
Nic and Murray Scarce came up with the idea when relatives were visiting Australia from Britain.
"During one of her trips to Australia, my mother-in-law acted as a magnet for all of our country’s big-name snakes, spiders and insects," Murray said in a blog post. "The questions relating to their identification and danger levels were relentless, and the fact that we didn’t have all the answers simply exacerbated the situation."
"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," said Critterpedia's project lead, Dr Matt Adcock.
The team behind the app understands very well how much info is needed and how important it is to get it all right. "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," Adcock said.
Every image that gets uploaded serves to help the machine learning algorithm, well, learn.
The app hasn't been released to the public just yet as it's still in beta testing but its creators have high hopes that when it's out there, it will help humans and animals get along.
"Critterpedia can create a world where people of all ages, backgrounds and status can appreciate and respect our environment, and where we and animals can peacefully coexist," Murray said.
h/t: Algorithm
