Lexical Emphasis Detection in Spoken French using F-BANKs and neural networks

Abdelwahab Heba (1, 2) , Thomas Pellegrini (1) , Tom Jorquera (2) , Régine André-Obrecht (1, 3) , Jean-Pierre Lorré (2)

1 IRIT-SAMoVA - Équipe Structuration, Analyse et MOdélisation de documents Vidéo et Audio
2 Linagora Labs [Toulouse]
3 UT3 - Université Toulouse III - Paul Sabatier

Expressiveness and non-verbal information in speech are active research topics in speech processing. In this work, we are interested in detecting emphasis at word-level as a mean to identify what are the focus words in a given utterance. We compare several machine learning techniques (Linear Discriminant Analysis, Support Vector Machines, Neural Networks) for this task carried out on SIWIS, a French speech synthesis database. Our approach consists first in aligning the spoken words to the speech signal and second to feed classifier with filter bank coefficients in order to take a binary decision at word-level: neutral/emphasized. Evaluation results show that a three-layer neural network performed best with a 93% accuracy.

Mots clés: Non verbal information in speech; SIWIS French speech synthesis database; Emphasized content recognition

Domaines: Informatique et langage [cs.CL]

Abdelwahab Heba, Thomas Pellegrini, Tom Jorquera, Régine André-Obrecht, Jean-Pierre Lorré. Lexical Emphasis Detection in Spoken French using F-BANKs and neural networks. International Conference on Statistical Language and Speech Processing (SLSP 2017), Oct 2017, Le Mans, France. pp.241-249, ⟨10.1007/978-3-319-68456-7_20⟩. ⟨hal-02559768⟩