Predicting Birdsongs Using Machine Learning

Scientists at the University of California, San Diego, ​have created a brain-to-tweet interface that is able to predict birdsongs before they are sung. 

This interface is able to decode the complex communication signals from a zebra finch's neural activity and reproduce realistic synthetic birdsong. The two song bites to the left represent the birdsong from the zebra zinc (top), and the predicted birdsong from the interface (bottom). 

What makes this accomplishment significant is its implications for human application. Birdsongs are complex in their pitch range and their vocalizations are similar to human speech. The experiment used machine learning to accomplish its goal and represents a significant step towards human brain reading device capabilities. 

Click here to read the full article.