AI Tracks 34 of 39 Wildlife Species in Malaysian Forests Using Soundwave Analysis
Karmactive Staff
In tropical woods, animal sounds are converted into spectrograms using AI algorithms to identify species.
34 of the 39 species under observation have been successfully identified by researchers at the University of Copenhagen.
Long stretches of continuous animal vocalizations are recorded by autonomous recording devices.
Machine learning algorithms interpret raw audio data, turning sound patterns into visual spectrograms.
Mammals, birds, frogs, and insects can all be identified by using AI algorithms that have been trained on annotated wildlife recordings.
The system is trained on annotated recordings, allowing it to swiftly adapt to novel species.
Where standard cameras fall short, AI-based sound monitoring aids in tracking elusive species in aquatic and dense forest environments.
Already, endangered species like Danish grey seals and Hawaiian monk seals are being monitored with this technique.
Amidst the ongoing mass extinction, AI-assisted bioacoustics technology assists long-term biodiversity conservation efforts.