Practical Guidelines for the use of AVDP

Tech Report 032

A practical guidelines document for the use of the Audiovisual Description Profile (MPEG-/ AVDP) for automatic metadata extraction purposes.

Open file (pdf, 3.2 MB)

Email me a link

Metadata arrives from a variety of sources: production systems, manual annotations and automatic annotations. As manual annotation is very time consuming and expensive, many archives lack the resources to perform this systematically, and use tools for the automatic extraction of spatiotemporal metadata. Time based metadata is metadata that is associated with a specific time code related to the audiovisual material.

MPEG-7 – a metadata standard created by MPEG - has the capability for representing results of automatic analysis tools using specific descriptors for low-level features (e.g., colour, shape, texture) which are quite useful in many automated information extraction applications. To improve interoperability with other metadata environments such as EBUCore, the EBU has developed a common metadata format for automated information extraction applications that adopts a subset of MPEG-7. The name of this format is MPEG-7 AVDP (AudioVisual Description Profile). It is formally a profile of MPEG-7 and it became an ISO/IEC international standard (ISO/IEC 15938 Part 9/AMD-1) in 2012.

Technical Report TR 032 provides practical guidelines for the use of this Audio Visual Description Profile (AVDP).

Translations

The following translations of this document are available. Please note the translations may not be based on the latest version; see the year mentioned in brackets. These translations are courtesy of the translaters and hosted by the EBU 'as is'. The official and definitive version is the EBU English version.

Russian version of EBU Tech Report 032 [2015] - Special thanks to EBU Member RTR.