Moving Picture, Audio, and Data Coding by Artificial Intelligence

New audio data coding standards body proposed

(Source: Depositphotos)

Leonardo Chiariglione, an Italian engineer best known as the co-founder and former chairman of MPEG, is proposing to establish a new organization called Moving Picture, Audio, and Data Coding by Artificial Intelligence (MPAI). The organization will have a twofold mission:

  1. to develop standards for compression of video, audio, and data, and
  2. to bridge the gap between its standards and their practical use.

This will be achieved by extending the MPEG process adding, before the start of the technical work (i.e., before issuing a Call for Proposals), the definition of a “Framework License.” The FWL is essentially the business model that IPR holders agree to adopt to remunerate their IPRs in the standard. However, the FWL does not contain “values” (e.g., royalty amounts, percentages, dates, etc.).

Before the start of the technical work, active members (i.e., those who will participate in the technical development) will:

  1. develop and adopt the FWL by a qualified majority
  2. individually declare that they will
  3. make available, within one year after approval of the standard, a license of their essential patents drafted according to the FWL.
  4. take a license of other members’ essential patents, if used, within one year after the licenses have been made available by the other IPR holders.

MPAI Founding Members will have an ongoing privilege to appoint about 1/3 of the directors of the board. You can see more details at Please note the link to the draft Statutes of MPAI at

There are related areas of study. For instance, the Khronos Organization is also looking into machine learning; however, they are not (currently) investigating compression using AI, which is the focus of MPAI. SMPTE also uses artificial intelligence but their purview is production and that is not what MPAI is targeting.

MPAI’s intention is to update the MPEG business model and make it thrive again in MPAI.