Towards automatic annotation of prosodic prominence levels in Austrian German

Julian Linke, Anneliese Kelterer, Markus Alexander Dabrowski, Dina El Zarka, Barbara Schuppler

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem Konferenzband


The creation of prosodic annotations is one of the most diffi-cult and time-consuming aspects of creating a speech database. Generally, only the speech signal and manually created tran-scriptions are available in an early resource development stage.This paper presents a tool for annotating prosodic prominenceat the word level, using exclusively acoustic features (96 f0-,intensity- and durational features). The best performance forseparating prominent from non-prominent words in Austrianread speech was reached with a decision tree with the abso-lute word duration as the only feature. For distinguishing moreprominence levels, a good performance was reached with a ran-dom forest model, similar to the best inter-annotator agreement.Furthermore, we analyzed in detail the feature ranking of therandom forest to give us insights into the relative importanceof the features contributing to prominence in Austrian German:Word duration>f0 range, RMS range. The specific findingsof this study will mainly be relevant for speech scientists andprosody researchers interested in German. Our methodologi-cal approach of analyzing prosodic prominence from a purelyacoustic perspective at the word-level will also be interestingfor researchers focusing on prosody in other languages.
TitelProceedings of Speech Prosody 2020
Untertitel10th International Conference on Speech Prosody 2020
ErscheinungsortTokyo, Japan
Seiten1000 - 1004
PublikationsstatusVeröffentlicht - 2020
Veranstaltung10th International Conference on Speech Prosody - virtuell, Keine Angaben
Dauer: 24 Mai 202028 Mai 2020


Konferenz10th International Conference on Speech Prosody
KurztitelSpeech Prosody 2020
LandKeine Angaben

Fingerprint Untersuchen Sie die Forschungsthemen von „Towards automatic annotation of prosodic prominence levels in Austrian German“. Zusammen bilden sie einen einzigartigen Fingerprint.

  • Projekte


    Elise Richter

    Barbara Schuppler (Empfänger/-in), Nov 2017

    Auszeichnung: Forschungsstipendium


    Wie der Mensch von der Maschine lernt

    Barbara Schuppler & Gernot Kubin


    1 eigener Medienbeitrag

    Presse/Medien: Presse / Medien

    Dieses zitieren

    Linke, J., Kelterer, A., Dabrowski, M. A., El Zarka, D., & Schuppler, B. (2020). Towards automatic annotation of prosodic prominence levels in Austrian German. in Proceedings of Speech Prosody 2020: 10th International Conference on Speech Prosody 2020 (S. 1000 - 1004). Tokyo, Japan.