Las Vegas, Nevada USA  /  April 22, 2017  -  April 27, 2017

NAB Show 2017

At NAB Show 2017 in Las Vegas, Fraunhofer IDMT presents its latest technologies for A/V analysis and search at Fraunhofer-booth in the South Hall (upper), Stand SU6110.


Audio and Video Segment Matching

Presentations include Audio Segment Matching and Video Segment Matching technologies, which detect complete/partial matches among files or live streams. They can be used

  • to detect complete and partial duplicates in A/V archives
  • to automatically track and propagate metadata for material reuse during content production and within content archives
  • to detect content reuse, which can be a powerful tool for program analysis and content monitoringg

Automatic A/V Annotation

Building on its existing A/V analysis technologies related to:

  • Music and Speech Detection
  • Music Annotation and Similarity
  • Video Shot Detection
  • Actor Recognition
  • Video Motion Analysis

Fraunhofer IDMT now provides several tools for Audio Forensics that can be used to automatically detect codec types and parameters of previous processing steps (codec analysis), identify the devices used for recordings (microphone classification), to automatically detect edits (editing detection), as well as to identify an original version among a set of duplicates (phylogeny analysis). All of these tools can be used to support content production and metadata annotation with the aim to prevent quality problems.

MICO: An Open-Source Framework for multi-modal Analysis

The possibility to perform complex multi-modal content analysis is crucial to support advanced search and recommendation functionalities and to improve the quality of metadata. MICO is an open-source framework that addresses some of these challenges.

Automatic Quality Control (QC)

Finally, Fraunhofer IDMT will present the integration of its QC components for automated Quality Control of DCP/IMF content into the professional movie post-production solution easyDCP by Fraunhofer IIS.