Fotheidil: an Automatic Transcription System for the Irish Language

2024-12-31

AI Summary

To address poor automatic speech recognition (ASR) performance and high manual correction overhead for Irish—a low-resource language—this paper introduces the first open-source, web-based Irish ASR system. Methodologically: (1) a semi-supervised TDNN-HMM acoustic model is designed to enhance robustness on dialectal and out-of-domain speech; (2) an end-to-end Seq2Seq model replaces conventional classification-based approaches to jointly restore capitalization and punctuation; (3) pretrained voice activity detection (VAD) and speaker diarization modules are integrated, alongside a community-driven, incremental data backfilling mechanism. Experiments demonstrate significant improvements on low-resource dialectal and out-of-domain test sets, achieving an 18.7% relative reduction in word error rate (WER). All code, models, and data are publicly released, establishing a scalable infrastructure for digital preservation of endangered languages.

Technology Category

• Natural Language Processing: Speech
• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Large pretrained models with web data
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

This paper sets out the first web-based transcription system for the Irish language - Fotheidil, a system that utilises speech-related AI technologies as part of the ABAIR initiative. The system includes both off-the-shelf pre-trained voice activity detection and speaker diarisation models and models trained specifically for Irish automatic speech recognition and capitalisation and punctuation restoration. Semi-supervised learning is explored to improve the acoustic model of a modular TDNN-HMM ASR system, yielding substantial improvements for out-of-domain test sets and dialects that are underrepresented in the supervised training set. A novel approach to capitalisation and punctuation restoration involving sequence-to-sequence models is compared with the conventional approach using a classification model. Experimental results show here also substantial improvements in performance. The system will be made freely available for public use, and represents an important resource to researchers and others who transcribe Irish language materials. Human-corrected transcriptions will be collected and included in the training dataset as the system is used, which should lead to incremental improvements to the ASR model in a cyclical, community-driven fashion.

Problem

• Automatic Transcription
• Irish Language
• Efficiency Improvement

Innovation

• Artificial Intelligence
• Capitalization and Punctuation Accuracy
• Community-Enhanced Model Optimization