AI might have finally decoded singing of birds
What's the story
In a groundbreaking study, scientists have used artificial intelligence (AI) to analyze over 116,000 bird songs from more than 3,000 species. The research found that despite their diversity, these birdsongs are built from just eight basic acoustic building blocks. The findings could revolutionize our understanding of avian communication and evolution.
Research methodology
How AI categorized the songs
The research team employed AI to analyze recordings from various public databases, including Xeno-canto.
The algorithm identified recurring sound patterns in over 3,000 species of songbirds, making it one of the largest studies ever conducted on this topic.
The AI categorized birdsongs into eight basic sound groups: slow trills, fast trills, ultrafast trills, flat whistles, slow-modulated whistles, fast-modulated whistles, harmonic stacks, and chaotic notes.
Song composition
Birds use limited 'sound parts' to create songs
Birds use these basic sound parts in different combinations to create their distinct songs, much like humans form words from a limited set of letters.
The study also found that birds living in dense forests tend to sing simpler songs that can penetrate the foliage cloud better.
In contrast, species in open areas can afford slower, more elaborate songs due to less acoustic interference.
Factors affecting song
Other factors influencing birdsong
The study also found that a bird's body size, beak shape, mating habits, and territory can affect its singing style.
This highlights the potential of AI to uncover patterns in nature that were previously difficult to detect.
The research also sheds light on why birds living in forests sing differently from those in open landscapes.
Communication
Why do birds sing?
Birds use their songs for a variety of reasons, including attracting mates, defending territory, and communicating with rivals or neighbors.
The study treated birdsongs as information rather than mere sound, analyzing how both frequency and timing change throughout each song.
This allowed the researchers to compare songs from thousands of species using the same method.