Tagarela - A Portuguese speech dataset from podcasts

TL;DR

The paper introduces TAGARELA, a large-scale Portuguese speech dataset from podcasts, to support the development of ASR and TTS models, addressing resource scarcity in Portuguese speech processing.

Contribution

It provides a publicly available, high-quality, large-scale Portuguese speech dataset with validated transcriptions, enabling improved speech technology development for Portuguese.

Findings

ASR and TTS models trained on TAGARELA perform effectively

Dataset quality validated through model-based transcription accuracy

Public release to foster Portuguese speech technology research

Abstract

Despite significant advances in speech processing, Portuguese remains under-resourced due to the scarcity of public, large-scale, and high-quality datasets. To address this gap, we present a new dataset, named TAGARELA, composed of over 8,972 hours of podcast audio, specifically curated for training automatic speech recognition (ASR) and text-to-speech (TTS) models. Notably, its scale rivals English's GigaSpeech (10kh), enabling state-of-the-art Portuguese models. To ensure data quality, the corpus was subjected to an audio pre-processing pipeline and subsequently transcribed using a mixed strategy: we applied ASR models that were previously trained on high-fidelity transcriptions generated by proprietary APIs, ensuring a high level of initial accuracy. Finally, to validate the effectiveness of this new resource, we present ASR and TTS models trained exclusively on our dataset and evaluate their performance, demonstrating its potential to drive the development of more robust and natural speech technologies for Portuguese. The dataset is released publicly, available at https://freds0.github.io/TAGARELA/, to foster the development of robust speech technologies.