Thaseus

THASEUS is an advanced analyzing system for audio and video content. It combines artificial intelligence, speech processing, and natural linguistics into one powerful framework. Running on state-of-the art speech-to-text algorithms, THASEUS offers an automatic generation of meta data in real time and stunning quality.

As core technology for mediaclipping's web portal, THASEUS is being developed and constantly improved by our R & D team. Key aspects of this advanced speech-to-text engine are:

  • Real-time processing and decoding
  • Language-independent engine: German and English is currently supported,
    French and Spanish will be added soon
  • >150.000 trained words
  • Speaker-, gender-, and dialect independent models
  • High-speed topic categorization and automatic summarization
  • Output: XML-based meta data

The facts sheet offers further technical information about Thaseus:
Thaseus Facts Sheet (pdf)