公开(公告)号：US10977555B2

公开(公告)日：2021-04-13

申请号：US16521756

申请日：2019-07-25

Applicant: Spotify AB

Inventor： Andreas Simon Thore Jansson ,  Angus William Sackfield ,  Ching Chuan Sung ,  Rachel M. Bittner

IPC: G06F15/00 ,  G10L25/00 ,  G06N3/08 ,  G06K9/62 ,  G06N3/04 ,  G10H1/00

Abstract: A system, method and computer product for training a neural network system. The method comprises inputting an audio signal to the system to generate plural outputs f(X, Θ). The audio signal includes one or more of vocal content and/or musical instrument content, and each output f(X, Θ) corresponds to a respective one of the different content types. The method also comprises comparing individual outputs f(X, Θ) of the neural network system to corresponding target signals. For each compared output f(X, Θ), at least one parameter of the system is adjusted to reduce a result of the comparing performed for the output f(X, Θ), to train the system to estimate the different content types. In one example embodiment, the system comprises a U-Net architecture. After training, the system can estimate various different types of vocal and/or instrument components of an audio signal, depending on which type of component(s) the system is trained to estimate.