Search Results for author:
Found 16 papers, 1 papers with code
Using a Knowledge Base to Automatically Annotate Speech Corpora and to Identify Sociolinguistic Variation
Speech characteristics vary from speaker to speaker.
Extracting Linguistic Knowledge from Speech: A Study of Stop Realization in 5 Romance Languages
This paper builds upon recent work in leveraging the corpora and tools originally used to develop speech technologies for corpus-based linguistic studies.
Multiscale Contextual Learning for Speech Emotion Recognition in Emergency Call Center Conversations
This paper presents a multi-scale conversational context learning approach for speech emotion recognition, which takes advantage of this hypothesis.
Exploring Attention Mechanisms for Multimodal Emotion Recognition in an Emergency Call Center Corpus
Two pre-trained models based on speech and text were fine-tuned for speech emotion recognition.
End-to-End Speech Emotion Recognition: Challenges of Real-Life Emergency Call Centers Data Recordings
Using the same end-to-end deep learning architecture, an Unweighted Accuracy Recall (UA) of 63% is obtained on IEMOCAP and a UA of 45. 6% on CEMO, each with 4 classes.
L\'enition et fortition des occlusives en coda finale dans deux langues romanes : le fran\ccais et le roumain (Lenition and fortition of word-final stops in two Romance languages: French and Romanian)
D{\`e}s lors, il est int{\'e}ressant de se pencher sur des ph{\'e}nom{\`e}nes phonologiques largement attest{\'e}s dans les langues en diachronie comme en synchronie pour {\'e}tablir leur {\'e}mergence ou non dans des langues qui n{'}y sont pas encore sujettes.
Lenition and Fortition of Stop Codas in Romanian
The present paper aims at providing a first study of lenition- and fortition-type phenomena in coda position in Romanian, a language that can be considered as less-resourced.
R\'ealisation phon\'etique et contraste phonologique marginal : une \'etude automatique des voyelles du roumain (Phonetic realization and marginal phonemic contrast : an automatic study of the Romanian vowels)
Les objectifs sont : (1) d{\'e}crire les traits acoustiques des voyelles en fonction du style de parole ; (2) estimer la relation entre traits acoustiques et contrastes phon{\'e}miques de la langue ; (3) estimer dans quelle mesure l{'}{\'e}tude de l{'}oral apporte des {\'e}clairages au sujet des attributs phon{\'e}miques des voyelles centrales [2] et [1], dont le statut (phon{\`e}mes vs allophones) est controvers{\'e}.
Automatic language identity tagging on word and sentence-level in multilingual text sources: a case-study on Luxembourgish
This language identification system was used to select textual data extracted from the web, in order to build a lexicon and language models.
Automatic Speech Recognition
Automatic Speech Recognition (ASR)
+4
Human annotation of ASR error regions: Is ``gravity'' a sharable concept for human annotators?
This paper is concerned with human assessments of the severity of errors in ASR outputs.
Cross-lingual studies of ASR errors: paradigms for perceptual evaluations
It is well-known that human listeners significantly outperform machines when it comes to transcribing speech.
Automatic Speech Recognition
Automatic Speech Recognition (ASR)
+2
