A recent review takes a close look at the rise of passive acoustic monitoring (PAM) in biological and ecological studies. With the increasing use of PAM comes a flood of data that researchers are struggling to keep up with. Manually sorting through this information is becoming less practical, especially as datasets grow larger and more complex.
Fortunately, advancements in machine learning offer a potential fix. Automated detection tools are being developed to help sift through these huge audio datasets, flagging important patterns and sounds. But despite the progress, the review notes that the use of these tools is still in its early stages. A key challenge is the knowledge gap between biologists and computer scientists, two fields that don’t always speak the same language.
The review aims to address this divide by offering a practical guide for biologists interested in using machine learning for bioacoustics. It also introduces key machine learning concepts for those unfamiliar with the tech side.