Audio Reuse Detection
Detect perceptually identical audio segments/matches among audio items or streams.
Fraunhofer IDMT
Developed by
License
Other
Intellectual property of Fraunhofer IDMT (closed source)
Main Characteristic
Audio reuse detection is a technology that detects perceptually identical audio segments/matches within a set of audio items or within audio streams. It can be used for duplicate detection and removal, metadata and rights tracking within and across archives, and to detect content reuse e. g. for media analysis.
Last updated
24.01.2023 - 11:22
Detailed Description
Audio reuse detection can be used to
- detect partial duplicates created by reuse during the production process, in order to flag or remove them
- detect partial reuse within a production system in order to track rights for a specific production
- detect manipulations based on reusing pre-existing audio material, in order to support content verification
- propagate metadata, by allowing participating systems or users to establish a system where partial duplicates are detected and respective metadata can be exchanged, reused and curated collaboratively, but without the need to access the actual content; the same approach can also be applied to ensure a proposer selection of training material for collaborative training of AI models
- detect reuse of material across radio / TV stations and over time, which can be used for media analysis, e. g. to find out how news reports to a certain topic were presented over time and/or across different broadcasters.