TEDxTN: A Three-way Speech Translation Corpus for Code-Switched Tunisian Arabic - English

TL;DR

This work addresses data scarcity for Tunisian Arabic code-switching speech translation by releasing TEDxTN, the first public Tunisian Arabic–English CS ST corpus. It collects 108 TEDx talks totaling approximately 25 hours, annotates them with CODA Habash guidelines, and provides robust ASR and AST baselines through domain-adapted pre-trained models in a SpeechBrain framework. Key findings show strong domain-adapted ASR performance with a best WER of 21.37% and competitive translation quality with BLEU scores around 25.7 after fine-tuning on TEDxTN data, alongside comprehensive CS statistics and trigger word analyses. By open-sourcing data, guidelines, and recipes, the work enables reproducibility and accelerates Tunisian dialect CS NLP research and extension to new talks and tasks.

Abstract

In this paper, we introduce TEDxTN, the first publicly available Tunisian Arabic to English speech translation dataset. This work is in line with the ongoing effort to mitigate the data scarcity obstacle for a number of Arabic dialects. We collected, segmented, transcribed and translated 108 TEDx talks following our internally developed annotations guidelines. The collected talks represent 25 hours of speech with code-switching that cover speakers with various accents from over 11 different regions of Tunisia. We make the annotation guidelines and corpus publicly available. This will enable the extension of TEDxTN to new talks as they become available. We also report results for strong baseline systems of Speech Recognition and Speech Translation using multiple pre-trained and fine-tuned end-to-end models. This corpus is the first open source and publicly available speech translation corpus of Code-Switching Tunisian dialect. We believe that this is a valuable resource that can motivate and facilitate further research on the natural language processing of Tunisian Dialect.