The MPEG-7 AudioVisual Description Profile (AVDP) has become a new ISO International Standard. The profile is designed for representing and exchanging metadata generated by automatic extraction tools, such as speech-to-text transcription engines and videoshot detection algorithms.
Sketch of the basic structure of AVDP descriptions ( source ).
According to EBU's Metadata Expert Jean-Pierre Evain, the new profile's standardization "marks a significant step towards making metadata cheaper through the processing of audio and video with minimal human intervention."
The AVDP concept was born in the EBU MIM/SCAIE project in 2010. Actively supported and promoted by RAI R&D and the VRT, the NHK and EBU have accompanied the proposal through the ISO standardization process. They were supported by industrial and research institutions JOANNEUM Research (Austria) and AT&T (USA).
Fits well with FIMS
AVDP has been designed to benefit the EBU community and the wider audiovisual industry, but it is not the end of the story. The EBU project MIM-SCAIE and the FP7 European project TOSCA-MP are working on guidelines and solutions for integrating extraction tools in production workflows. Use cases have been submitted to the EBU-AMWA Framework for Interoperable Media Services (FIMS). The FIMS project is developing standardized Service Oriented Architectures (SOA) for media applications. It will now be able to define associated services using AVDP as a standardised common metadata format. For more information contact: Jean-Pierre Evain (EBU) .