Enhancing Standard and Dialectal Frisian ASR: Multilingual Fine-tuning and Language Identification for Improved Low-resource Performance
We improve Frisian ASR performance by using multilingual fine-tuning and language identification, highlighting the challenges in dialectal speech recognition and the importance of elicitation methods in evaluating real-world performance.
Mar 7, 2025
Evaluating Standard and Dialectal Frisian ASR: Multilingual Fine-tuning and Language Identification for Improved Low-resource Performance
We improve ASR for Frisian and its dialects through multilingual fine-tuning and auxiliary language identification, while highlighting the challenges of dialectal speech recognition.
Feb 7, 2025
Fine-Tuning Strategies for Dutch Dysarthric Speech Recognition: Evaluating the Impact of Healthy, Disease-Specific, and Speaker-Specific Data
We compare three fine-tuning strategies for Dutch dysarthric speech recognition, finding that combining healthy and dysarthric speech, as well as speaker-specific approaches, yield the best results.
Sep 1, 2024
A Functional Trade-off between Prosodic and Semantic Cues in Conveying Sarcasm
We explore how speakers balance prosodic and semantic cues when expressing sarcasm, finding that prosodic features become more pronounced when semantic markers are ambiguous.
Aug 1, 2024
Meta Learning Text-to-Speech Synthesis in over 7000 Languages
We developed a meta-learning approach for text-to-speech synthesis that works across 7000+ languages, especially benefiting under-resourced and endangered languages.
Jun 1, 2024
SarcasticSpeech: Speech Synthesis for Sarcasm in Low-Resource Scenarios
We present a first attempt at sarcastic speech synthesis using transfer learning to overcome the challenge of limited sarcastic speech data.
Aug 1, 2023
Strategies in Transfer Learning for Low-Resource Speech Synthesis: Phone Mapping, Features Input, and Source Language Selection
A comparative study of different transfer learning strategies for text-to-speech synthesis in low-resource languages.
Jun 21, 2023
The Effects of Input Type and Pronunciation Dictionary Usage in Transfer Learning for Low-Resource Text-to-Speech
A study of input types and pronunciation dictionary alternatives for text-to-speech in low-resource languages, with a focus on West Frisian.
Jun 1, 2023
Resource-Efficient Fine-Tuning Strategies for Automatic MOS Prediction in Text-to-Speech for Low-Resource Languages
Research on efficient strategies for automatic Mean Opinion Score prediction in text-to-speech for low-resource languages, with a case study on West Frisian.
May 30, 2023
Improving Luxembourgish Speech Recognition with Cross-Lingual Speech Representations
A study on improving automatic speech recognition for Luxembourgish through cross-lingual speech representations and transfer learning.
Jan 1, 2023