No. of Recommendations: 9
The goal was to identify birds by their song.
"Automatic song ID has been a dream for decades, but analyzing sound has always been extremely difficult."
I guess there's a reason why animation uses drawings for the visual part but voice actors for the audio part (contemplate for a moment the inverse: a live-action movie dubbed only by artificial voices.)
Then a great insight:
"The breakthrough came when researchers began treating the sounds as images and applying new and powerful image classification algorithms like the ones that power Merlin’s Photo ID feature. The Merlin team was led by Grant Van Horn, who is now an assistant professor at University of Massachusetts Amherst.
“Each sound recording a user makes gets converted from a waveform to a spectrogram—a way to visualize the amplitude [volume], frequency [pitch], and duration of the sound,” Van Horn says. “So just like Merlin can identify a picture of a bird, it can now use this picture of a bird’s sound to make an ID,” Van Horn says."
...and, it works beautifully. I step outside with my morning coffee and the (free) app says yes, those are red-winged blackbirds, Bullock's orioles, acorn woodpeckers...even accurately identifying the honk of the nesting great blue herons across the field.
Impressive, insightful, nondestructive, non-exploitative.
Sometimes humans can still get it right.
--sutton
