Algorithm almost im-peck-able at picking up chicken distress calls

Publicly released:
International

International researchers have developed a deep learning tool to identify the distress calls of farmed chickens. With over 25 billion chickens farmed each year, monitoring chicken noises for distress could be an easy way to ensure there's no fowl play on farms, and that chickies get the help they need, when they need it. Using recordings from an intensive farm, the team developed an algorithm that correctly identified 97% of distress calls, among other farm noises. This research has eggs-taordinary potential for developing technologies to monitor distress calls in large, commercial chicken flocks and assess the welfare of our feathered friends around the world.

Media release

From: The Royal Society

Automated identification of chicken distress vocalisations using deep learning models

Annual global production of chickens exceeds 25 billion, and they are often housed in groups numbering thousands. Thus, assessing their welfare is very challenging. The amount of chicken distress calls could be used as an “iceberg indicator” of welfare because it’s linked to growth and mortality levels. We developed an algorithm to automatically identify chicken distress calls using recordings from an intensive farm. We correctly identified 97% of distress calls, among other farm sounds. This research will hopefully allow the development of precision livestock farming technology that will seek to monitor and reduce the stress experienced by chickens during production.

Multimedia

distressed-chickens_RSI_audio.wav
distressed-chickens_RSI_audio.wav
Journal/
conference:
Journal of the Royal Society Interface
Research:Paper
Organisation/s: City University of Hong Kong, China
Funder: The research was carried out as part of the LIVEQuest project supported by InnovateUK and BBSRC grant no. 2016YFE01242200.
Media Contact/s
Contact details are only visible to registered journalists.