Evaluation of recognition errors of hybrid and transformer-based ASR systems in German video lectures

  • We analyze different errors in speech recognition systems, focusing on consecutive insertions and deletions, known as hallucinations and elisions in transformer-based end-to-end automatic speech recognition (ASR) systems. We compare errors from a TDNN-HMM, and whisper-based models on English and German spontaneous speech. Based on a human annotated subset of German lecture videos, we investigate whether these blocks of deletions affect the semantics of the utterance. Whisper performs best and preserves the meaning in 90% of the annotated error segments even containing consecutive deletions on this subset. We analyze the word error rate and do further analysis of errors using natural language processing to detect lemmatization errors, compound word errors, and out-of-vocabulary words. We discuss possible reasons and mitigations.

Export metadata

Additional Services

Search Google Scholar
Metadaten
Author:Thomas Ranzenberger, Ilja Baumann, Sebastian Bayerl, Dominik Wagner, Tobias BockletORCiD, Korbinian Riedhammer
URL / DOI:https://www.essv.de/pdf/2025_101_108.pdf
Parent Title (German):Studientexte zur Sprachkommunikation: Elektronische Sprachsignalverarbeitung 2025 - Book
Publisher:ESSV 2025
Place of publication:Halle, Deutschland
Document Type:Article
Language:English
Date of first Publication:2025/03/05
Reviewed:Begutachtet/Reviewed
Release Date:2025/11/06
Pagenumber:101-108
Konferenzangabe:ESSV 2025
institutes:Fakultät Informatik
Zentrum für Künstliche Intelligenz (KIZ)
Research Themes:Digitalisierung & Künstliche Intelligenz
Verstanden ✔
Diese Webseite verwendet technisch erforderliche Session-Cookies. Durch die weitere Nutzung der Webseite stimmen Sie diesem zu. Unsere Datenschutzerklärung finden Sie hier.