Search Results for author: Damien Lolive

Found 20 papers, 0 papers with code

A Low-Cost Motion Capture Corpus in French Sign Language for Interpreting Iconicity and Spatial Referencing Mechanisms

no code implementations LREC 2022 Clémence Mertz, Vincent Barreaud, Thibaut Le Naour, Damien Lolive, Sylvie Gibet

The automatic translation of sign language videos into transcribed texts is rarely approached in its whole, as it implies to finely model the grammatical mechanisms that govern these languages.

Investigating Inter- and Intra-speaker Voice Conversion using Audiobooks

no code implementations LREC 2022 Aghilas Sini, Damien Lolive, Nelly Barbot, Pierre Alain

Audiobook readers play with their voices to emphasize some text passages, highlight discourse changes or significant events, or in order to make listening easier and entertaining.

Speech Synthesis Text-To-Speech Synthesis +1

Neural-Driven Search-Based Paraphrase Generation

no code implementations EACL 2021 Betty Fabre, Tanguy Urvoy, Jonathan Chevelu, Damien Lolive

We study a search-based paraphrase generation scheme where candidate paraphrases are generated by iterated transformations from the original sentence and evaluated in terms of syntax quality, semantic distance, and lexical distance.

Paraphrase Generation

Style versus Content: A distinction without a (learnable) difference?

no code implementations COLING 2020 Somayeh Jafaritazehjani, Gw{\'e}nol{\'e} Lecorv{\'e}, Damien Lolive, John Kelleher

Due to the lack of parallel data for style transfer we employ a variety of adversarial encoder-decoder networks in our experiments.

Style Transfer

Neural-Driven Multi-criteria Tree Search for Paraphrase Generation

no code implementations NeurIPS Workshop LMCA 2020 Betty Fabre, Tanguy Urvoy, Jonathan Chevelu, Damien Lolive

A good paraphrase is semantically similar to the original sentence but it must be also well formed, and syntactically different to ensure diversity.

Paraphrase Generation

\'Evaluation objective de plongements pour la synth\`ese de parole guid\'ee par r\'eseaux de neurones (Objective evaluation of embeddings for speech synthesis guided by neural networks)

no code implementations JEPTALNRECITAL 2019 Antoine Perquin, Gw{\'e}nol{\'e} Lecorv{\'e}, Damien Lolive, Laurent Amsaleg

La qualit{\'e} des plongements est li{\'e}e {\`a} la t{\^a}che sp{\'e}cifique pour laquelle ils ont {\'e}t{\'e} entra{\^\i}n{\'e}s et l{'}{\'e}valuation de cette t{\^a}che peut {\^e}tre un proc{\'e}d{\'e} long et on{\'e}reux s{'}il y a besoin d{'}annotateurs humains.

Speech Synthesis

Ajout automatique de disfluences pour la synth\`ese de la parole spontan\'ee : formalisation et preuve de concept (Automatic disfluency insertion towards spontaneous TTS : formalization and proof of concept)

no code implementations JEPTALNRECITAL 2017 Raheel Qader, Gw{\'e}nol{\'e} Lecorv{\'e}, Damien Lolive, Pascale S{\'e}billot

Cet article pr{\'e}sente un travail exploratoire sur l{'}ajout automatique de disfluences, c{'}est-{\`a}-dire de pauses, de r{\'e}p{\'e}titions et de r{\'e}visions, dans les {\'e}nonc{\'e}s en entr{\'e}e d{'}un syst{\`e}me de synth{\`e}se de la parole.

Une p\'enalit\'e floue fond\'ee phonologiquement pour am\'eliorer la S\'election d'Unit\'e (A Phonologically Motivated Penalty To Improve Unit Selection)

no code implementations JEPTALNRECITAL 2016 David Guennec, Damien Lolive

Les syst{\`e}mes de synth{\`e}se par corpus reposent, sauf de rares exceptions, sur des co{\^u}ts cibles et des co{\^u}ts de concat{\'e}nation pour s{\'e}lectionner la meilleure s{\'e}quence d{'}unit{\'e}s. Le r{\^o}le du co{\^u}t de concat{\'e}nation est de s{'}assurer que l{'}assemblage de deux segments de parole ne causera l{'}apparition d{'}aucun artefact acoustique.

NER

Patrons Rythmiques et Genres Litt\'eraires en Synth\`ese de la Parole (How to improve rhythmic patterns according to literary genre in synthesized speech ⇤ )

no code implementations JEPTALNRECITAL 2016 Elisabeth Delais-Roussarie, Damien Lolive, Hiyon Yoo, David Guennec

Ces vingt derni{\`e}res ann{\'e}es, la qualit{\'e} de la parole synth{\'e}tique s{'}est am{\'e}lior{\'e}e gr{\^a}ce notamment {\`a} l{'}{\'e}mergence de nouvelles techniques comme la synth{\`e}se par corpus.

Se concentrer sur les diff\'erences : une m\'ethode d'\'evaluation subjective efficace pour la comparaison de syst\`emes de synth\`ese (Focus on differences : a subjective evaluation method to efficiently compare TTS systems * )

no code implementations JEPTALNRECITAL 2016 Jonathan Chevelu, Damien Lolive, S{\'e}bastien Le Maguer, David Guennec

Cette m{\'e}thode est appliqu{\'e}e sur un syst{\`e}me de synth{\`e}se de type HTS et un second par s{\'e}lection d{'}unit{\'e}s. La comparaison avec l{'}approche classique montre que cette m{\'e}thode r{\'e}v{\`e}le des {\'e}carts qui jusqu{'}alors n{'}{\'e}taient pas significatifs.

ROOTS: a toolkit for easy, fast and consistent processing of large sequential annotated data collections

no code implementations LREC 2014 Jonathan Chevelu, Gw{\'e}nol{\'e} Lecorv{\'e}, Damien Lolive

The development of new methods for given speech and natural language processing tasks usually consists in annotating large corpora of data before applying machine learning techniques to train models or to extract information.

Management

Towards Fully Automatic Annotation of Audio Books for TTS

no code implementations LREC 2012 Olivier Boeffard, Laure Charonnat, S{\'e}bastien Le Maguer, Damien Lolive

Nowadays, the use of statistical models implies the use of huge sized corpora that need to be recorded, transcribed, annotated and segmented to be usable.

Speech Recognition Speech Synthesis +1

Cannot find the paper you are looking for? You can Submit a new open access paper.