direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Inhalt des Dokuments

MPEG-7 SpokenContent Tool

Project information
Project Researcher:
Dr. Nicolas Moreau
Research Students:
Jin Shan, Martin Haller

The spoken content information extracted by automatic speech recognition (ASR) systems can be output in different ways and formats. It poses problems of compatibility for systems which need to access and exchange this information. The audio part of MPEG-7 (Part 4) contains a SpokenContent Description Scheme (DS) consisting of a standardized description of the information extracted by speech recognisers.

Online demonstrator

Online Demo : Upload a speech signal (in German) and get an MPEG-7 SpokenContent Descriptor.

Zusatzinformationen / Extras


Schnellnavigation zur Seite über Nummerneingabe