AI Can Predict Instrument Sounds
This research primarily focuses on the violin. The researchers used a feed-forward neural network with a single hidden layer and a sigmoid activation function connected to a linear output layer. The system is fed with 20 parameters of measurements to determine the shape of the outline. It then returns the eigenfrequency or modal frequency values of the first 10 vibrational modes of the violin plates. When put to test, the researchers observed that the system has an accuracy of almost 98%. This opens the possibility for violin makers to experiment with the materials for better efficiency before implementing them. However, it remains to be seen if a commercially viable service comes out of this research. The research paper explores the methods researchers used along with their findings in the process. If you’re interested to learn more about this, you can read the complete report from the report titled ‘A data-driven approach to violin making’ published in Nature journal right here. “The ability to predict how a violin design sounds, can truly be a game-changer for violin makers, as not only will it help them do better than the grand masters, but it will also help them explore the potential of new designs and materials,” wrote the researchers in the report.
