Listening to the bush: How AI can help NZ rid its wilderness of 'hold‑out' pest possums
Possums are among New Zealand's most destructive introduced mammals. They damage native forests, prey on wildlife and remain a major target of the country's Predator Free 2050 program.
Possums pose a significant threat to New Zealand's native forests and wildlife, as part of the country's Predator Free 2050 program. Identifying and removing the last few surviving possums can be challenging, as even a small number can repopulate an area. Microphones recording overnight, combined with AI software, show promise in detecting possums, but current AI models can produce false alarms by mistaking other animals for possums.
Researchers developed a new training approach called "cross-model confusion mapping," which used BirdNET, an AI trained to identify over 6,000 bird species, as a "hard negative" to teach the AI model which bird calls to distinguish from possum vocalizations. This method reduced false alarms without compromising detection accuracy, even when tested on new, unseen forest recordings. While the approach has shown promise, further field testing is needed before widespread deployment.
Written by urgent.news from Phys.org's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
