Digital audio processing is increasingly dependent on algorithms that learn various features and characteristics of audio signals. Neural networks are often used, and they have to be trained with large bodies of audio material so that they can start to behave in a predictable and useful way. Once trained they can be put to work in roles such as distinguishing between live and studio recordings, searching for specific drum sounds in mixes, creating morphed sounds, or emulating existing analog effects.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.