• search hit 21 of 301
Back to Result List

Cross-Lingual Acoustic Modeling in Upper Sorbian – Preliminary Study

  • In this paper, we present a preliminary study for acoustic modeling in Upper Sorbian, where a model of German was used in cross-lingual transfer learning. At first, we define the grapheme and phoneme inventories and map the target phonemes from the most similar German source equivalents. Phonetically balanced sentences for the recording prompts were selected from a combination of general and domain-specific textual data. The speech corpora with a total duration of around 11 hours was collected in controlled recording sessions involving an equal number of females, males, and children. The baseline acoustic model was employed to force-align the speech corpora given the knowledge-based phoneme mappings. How well the mappings were, was evaluated by the phoneme confusions in free-phoneme recognition. The new derived data-driven model with a reduced phoneme set was included in the adaptation and evaluation along with the baseline acoustic model. The model adaptation performance was cross-validated with the “Leave One Group Out” strategy. WeIn this paper, we present a preliminary study for acoustic modeling in Upper Sorbian, where a model of German was used in cross-lingual transfer learning. At first, we define the grapheme and phoneme inventories and map the target phonemes from the most similar German source equivalents. Phonetically balanced sentences for the recording prompts were selected from a combination of general and domain-specific textual data. The speech corpora with a total duration of around 11 hours was collected in controlled recording sessions involving an equal number of females, males, and children. The baseline acoustic model was employed to force-align the speech corpora given the knowledge-based phoneme mappings. How well the mappings were, was evaluated by the phoneme confusions in free-phoneme recognition. The new derived data-driven model with a reduced phoneme set was included in the adaptation and evaluation along with the baseline acoustic model. The model adaptation performance was cross-validated with the “Leave One Group Out” strategy. We observed major improvements in phoneme error rates after adaptation for the knowledge-based and data-driven phoneme mappings. The study confirmed the feasibility of transfer learning for acoustic model adaptation in the case of Upper Sorbian, at the same time demonstrating practical usability with a small vocabulary speech recognition application (Smart Lamp).show moreshow less

Export metadata

Additional Services

Search Google Scholar
Metadaten
Author: Ivan Kraljevski, Marek Rjelka, Frank Duckhorn, Constanze Tschöpe, Matthias WolffORCiD
URL:https://publica.fraunhofer.de/dokumente/N-633297.html
ISBN:978-3-959082-27-3
ISSN:0940-6832
Title of the source (German):Elektronische Sprachsignalverarbeitung 2021 : Tagungsband der 32. Konferenz Berlin, 3.-5. März 2021
Publisher:TUDpress
Place of publication:Dresden
Editor: Stefan Hillmann, Benjamin Weiss, Thilo Michael, Sebastian Möller
Document Type:Conference publication peer-reviewed
Language:German
Year of publication:2021
First Page:43
Last Page:50
Series ; volume number:Studientexte zur Sprachkommunikation ; Band 99
Faculty/Chair:Fakultät 1 MINT - Mathematik, Informatik, Physik, Elektro- und Informationstechnik / FG Kommunikationstechnik
Einverstanden ✔
Diese Webseite verwendet technisch erforderliche Session-Cookies. Durch die weitere Nutzung der Webseite stimmen Sie diesem zu. Unsere Datenschutzerklärung finden Sie hier.