One problem ecologists face when monitoring rare birds is that we don’t have enough recordings of them to utilize the various apps and software available for tracking.
To solve this, researchers are turning to deep learning AI to create a tool called ECOGEN that can mimic the songs of rare birds from small samples. It works by creating spectograms from recordings. These are visual representations of audio. New AI images are then generated from these. These new images are then converted back to audio that can continue to train bird sound identification models.
Whether it hallucinates, we don’t know, but it works. The researches say it’s improved song classification identification by 12 percent.
“Essentially, for species with limited wild recordings, such as those that are rare, elusive, or sensitive, you can expand your sound library without further disrupting the animals or conducting additional fieldwork,” Dr. Nicolas Lecomte, one of the project researchers, tells phys.org.
Here’s another something neat:
The ECOGEN tool has other potential applications. For instance, it could be used to help conserve extremely rare species, like the critically endangered regent honeyeaters, where young individuals are unable to learn their species’ songs because there aren’t enough adult birds to model them.
The tool could benefit other types of animal as well. Dr. Lecomte added, “While ECOGEN was developed for birds, we’re confident that it could be applied to mammals, fish (yes, they can produce sounds), insects and amphibians.”
Researchers and the curious can find the open source repository on Github.
