Content Analysis Technologies for Media Asset Management

The segment matching technology of Fraunhofer IDMT detects content reuse and helps to understand the processing history and relationships within a set of items.

IBC2018_Analysis

Content analysis technologies for media asset management

Fraunhofer IDMT will be presenting the latest versions of its automatic metadata extraction and fingerprinting technologies for media asset management, which can be integrated in existing systems and applications helping broadcasters and archives to fully exploit the potential of their content.

The following technologies will be demonstrated at IBC:

  • - A unique combination of A/V matching and phylogeny analysis tools to not only identify content reuse and overlaps, but also to
      understand its processing history and to distinguish between original content and transcoded content copies
  • - Various automatic metadata extraction technologies for the efficient annotation of A/V content , including e.g. speech and
      music detection, music annotation and similarity, video motion analysis, shot detection and object and actor recognition
  • - Automatic quality control tools for the detection of visual artefacts including blocking, blurring, ringing, noise, etc. and audio
      artefacts including clipping, dropout, phase shift, channel similarity or low-bitrate encoding
  • - Audio forensics tools to detect intended and unintended audio editing, and to support identification of audio fakes
  • - A technology for privacy-preserving data analysis and recommendation, which is based on a strong decoupling of real and
      virtual IDs in systems