Sound and music computing using AI: designing a standard

While there are currently various approaches that define and adapt the conditions in which the user experiences content or service for several music and audio-related applications including entertainment, communication, audio documents preservation/restoration, we are missing worldwide accepted standards that enable data exchange and interoperability based on common interfaces for such applications. The Moving Picture, Audio and Data Coding by Artificial Intelligence (MPAI) is an international non-profit organization whose mission is to develop such standards. Relying on Artificial Intelligence (AI), MPAI creates a workflow of AI Modules (AIM) that are interchangeable and upgradable without necessarily changing the logic of the application. A specific area of work, MPAI Context-based Audio Enhancement (MPAI-CAE), is showing tremendous possibilities for the Sound and Music Computing (SMC) community. MPAI-CAE applies context information to the input content to deliver the audio output via the most appropriate protocol. Three MPAI-CAE case studies particularly relevant for the SMC community will be presented in this paper: Audio recording preservation (ARP), a use case that covers the whole “philologically informed” archival process of an audio document, from the active sound documents preservation to the access to digitized files; Audio-on-the-go (AOG), which aims to improve safety and listening quality for situations in which the users are in motion in different environments; and Emotion-enhanced speech (EES), a use case that implements a user-friendly system control interface that generates speech with various levels of emotions.