Europarl-ST: A Multilingual Corpus for Speech Translation of Parliamentary Debates
By Javier Iranzo-Sánchez, Joan Albert Silvestre-Cerdà, Javier Jorge, Nahuel Roselló, Adrià Giménez, A. Sanchís, Jorge Civera Saiz, Alfons Juan-Císcar. 2019
216 citations recorded by the source corpus.
Current research into spoken language translation (SLT), or speech-to-text translation, is often hampered by the lack of specific data resources for this task, as currently available SLT datasets are restricted to a limited set of language pairs. In this paper we present Europarl-ST, a novel multilingual SLT corpus containing paired audio-text samples for SLT from and into 6 European languages, for a total of 30 different translation directions. This corpus has been compiled using the debates held in the European Parliament in the period between 2008 and 2012. This paper describes the corpus creation process and presents a series of automatic speech recognition, machine translation and spoken language translation experiments that highlight the potential of this new resource. The corpus is released under a Creative Commons license and is freely accessible and downloadable.
Subfields: Artificial Intelligence, Machine Translation, Speech & Audio
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Graham Neubig
- Luke S. Zettlemoyer
- Mohit Bansal
- Jiatao Gu
- Juan Pino
- Changhan Wang
- Kyunghyun Cho
- Furu Wei
In Machine Translation
In Speech & Audio
Citations in the corpus
- FLEURS: FEW-Shot Learning Evaluation of Universal Representations of Speech (2022, background): “While there are a few other datasets containing n-way parallel speech and text, including Europarl-ST [18], MuST-C [19], mTEDx [20] and the CVSS corpus [21], to the best of our knowledge, FLEURS is the only dataset spanning over 100 languages enabling research on a diverse set of languages and domai”
- VoxPopuli: A Large-Scale Multilingual Speech Corpus for Representation Learning, Semi-Supervised Learning and Interpretation (2021, methodology): “We train Transformer Base (Vaswani et al., 2017) supervised baselines and use 3K-hour monolingual VoxPopuli unlabelled data (from 2013-2020 sessions only to avoid overlaps with EuroParl-ST) to self-train Transformer Large models.”
- Enhanced Direct Speech-to-Speech Translation Using Self-supervised Pre-training and Data Augmentation (2022, methodology): “For VoxPopuli [53], we remove utterances from the year 2012 and before to avoid overlap with the Europarl-ST [36] dev and test data.”
- Multilingual Speech Translation from Efficient Finetuning of Pretrained Models (2021, methodology): “Previous SOTAs: We compare to the best end-to-end (E2E) model from previous literature (Wang et al., 2020b; Iranzo-S´anchez et al., 2020) on each translation direction, which is usually the best-performing multilingual model trained with parallel data from all directions (both X-En and En-X) and…”
- Pile of Law: Learning Responsible Data Filtering from the Law and a 256GB Open-Source Legal Dataset (2022, background): “Others have aggregated smaller subsets of legal data, such the EuroParl datasets which gather European Parliamentary debates [64, 57].”
- Speech Translation and the End-to-End Promise: Taking Stock of Where We Are (2020, background): “…of the latter approach, so follow-up work concentrates on compiling ST corpora from available web sources (Godard et al., 2018; Kocabiyikoglu et al., 2018; Sanabria et al., 2018; di Gangi et al., 2019a; Boito et al., 2020; Beilharz et al., 2020; Iranzo-Sánchez et al., 2020; Wang et al., 2020a).”
- SpeechMatrix: A Large-Scale Mined Corpus of Multilingual Speech-to-Speech Translations (2022, unknown): “It is also found that translation results are not symmetric for some language pairs, for example, ro-en has a BLEU of 22.6 while en-ro BLEU is only 7.6 on EPST.”
- CoVoST 2: A Massively Multilingual Speech-to-Text Translation Corpus (2020, background): “The development of benchmark datasets, such as MuST-C (Di Gangi et al., 2019), EuroparlST (Iranzo-Snchez et al., 2020) or CoVoST (Wang et al., 2020), has greatly contributed to the increasing popularity of speech-to-text translation (ST) as a research topic.”
- XTREME-S: Evaluating Cross-lingual Speech Representations (2022, background): “Europarl-ST [22], Must-C [23] and mTEDX [24] also provide common evaluation of speech translation.”
- The Multilingual TEDx Corpus for Speech Recognition and Translation (2021, methodology): “Finally, existing multi-target corpora do not have the same segmentations across language pairs [7, 11, 12], which complicates the analysis of multilingual models and language transfer.”
- Joint Pre-Training with Speech and Bilingual Text for Direct Speech to Speech Translation (2022, unknown): “(2) The proposed model achieves a significant improvement of about 5 BLEU scores compared to the pre-trained model without MT data.”
- CoVoST 2 and Massively Multilingual Speech-to-Text Translation (2020, unknown): “xtensive speech recognition, bilingual and multilingual machine translation and ST baselines. 1 Introduction The development of benchmark datasets, such as MuST-C (Di Gangi et al., 2019), EuroparlST (Iranzo-Snchez et al., 2020) or CoVoST (Wang et al., 2020), has greatly contributed to the increasing”
- FINDINGS OF THE IWSLT 2020 EVALUATION CAMPAIGN (2020, unknown): “…al., 2019a) • WIT3 (Cettolo et al., 2012) • Speech-Translation TED corpus7 • How2 (Sanabria et al., 2018)8 • LibriVoxDeEn (Beilharz and Sun, 2019)9 • Europarl-ST (Iranzo-Sánchez et al., 2020) • TED LIUM v2 (Rousseau et al., 2014) and v3 (Hernandez et al., 2018) • all the data provided by WMT…”
- Multimodal and Multilingual Embeddings for Large-Scale Speech Mining (2021, methodology): “, 2020a) provides speech translations in 15 En-X directions and 22 X-En directions, some languages are presented in Table 1, and EuroparlST (Iranzo-Sánchez et al., 2020) provides 30 different speech translation directions from 6 European languages.”