A new software program, developed by Department of Environment, Land, Water and Planning (DELWP) scientists, which identifies frog calls and sorts them from other sounds, could accelerate ecological research on frogs and other species.
DELWP said the team from its Arthur Rylah Institute for Environmental Research was behind the pilot Frog Call Recognition Project and found its software identified frogs in sound recordings far more accurately than existing systems.
“Researchers have turned to remote recording devices to save on the time and expense of sending ecologists into the field to listen for sounds like frog calls to identify the location of species,” DELWP said.
“But this creates a bottleneck of data – enormous numbers of sound recordings that are impossible for humans to listen to manually,” it said.
“The team got to talking about the problem of their data bottleneck.”
DELWP said the project was developed by Wildlife Ecologists at the Department, Katie Howard, Louise Durkin and Lachlan Francis; and Senior Scientist for Ecological Modelling at the Department, Peter Griffioen.
The Department said the team applied a deep learning approach known as convolutional neural networks, an Australian-first use of the technology in ecology.
“It’s similar to existing software like the Shazam song recognition app,” it said.
“A painstaking training process is needed, using samples of the sound files to ensure the software can identify frog calls and other sounds.”
DELWP said accuracy was critical to the success of the model and once the training phase was completed, the results of the pilot were “stunning”.
The Department said now the concept and technology had proved effective, there were plans to expand the project to birds and bats.
