• Publications
  • Influence
End-to-End Automatic Speech Translation of Audiobooks
Experimental results show that it is possible to train compact and efficient end-to-end speech translation models in this setup and hope that the speech translation baseline on this corpus will be challenged in the future. Expand
FlauBERT: Unsupervised Language Model Pre-training for French
This paper introduces and shares FlauBERT, a model learned on a very large and heterogeneous French corpus and applies it to diverse NLP tasks and shows that most of the time they outperform other pre-training approaches. Expand
Automatic speech recognition for under-resourced languages: A survey
This paper proposes, in this paper, a survey that focuses on automatic speech recognition (ASR) for under-resourced languages, and a literature review of the recent contributions made. Expand
Augmenting Librispeech with French Translations: A Multimodal Corpus for Direct Speech Translation Evaluation
This paper augments an existing (monolingual) corpus: LibriSpeech with an existing corpus derived from read audiobooks from the LibriVox project, and shows that the automatic alignments scores are reasonably correlated with the human judgments of the bilingual alignment quality. Expand
Step-by-step and integrated approaches in broadcast news speaker diarization
This paper summarizes the collaboration of the LIA and CLIPS laboratories on speaker diarization of broadcast news during the spring NIST Rich Transcription 2003 evaluation campaign (NIST-RTO03S).Expand
Listen and Translate: A Proof of Concept for End-to-End Speech-to-Text Translation
This paper proposes a first attempt to build an end-to-end speech- to-text translation system, which does not use source language text during learning or decoding, and would drastically change the data collection methodology in speech translation. Expand
Collection of a Large Database of French-English SMT Output Corrections
This paper describes the data collection and analysis of a large database of 10.881 SMT translation output hypotheses manually corrected and discusses the adequation of text-based generated reference translations to train setence-to-sentence based SMT systems. Expand
Automatic sound detection and recognition for noisy environment
The detection algorithm, based on a median filter, features a highly robust performance even under important background noise conditions, and a rather good recognition rate can be reached, even under severe gaussian white noise degradations. Expand
Breaking the Unwritten Language Barrier: The BULB Project
Abstract The project Breaking the Unwritten Language Barrier (BULB), which brings together linguists and computer scientists, aims at supporting linguists in documenting unwritten languages. In orderExpand
Localization and selection of speaker-specific information with statistical modeling
The authors propose a system that automatically selects the most discriminant parts of a speech utterance, which divides the signal into different time–frequency blocks and proposes a new selection procedure designed to select the pertinent features. Expand