Skip to Main content Skip to Navigation
Theses

Localisation binaurale active de sources sonores en robotique humanoïde

Alban Portello 1
1 LAAS-RAP - Équipe Robotique, Action et Perception
LAAS - Laboratoire d'analyse et d'architecture des systèmes
Abstract : This thesis deals with the definition of algorithms for the localisation of still or moving sound sources from a mobile binaural sensor in robotics. The aim is to develop active strategies, combining left-right perceived signals with the sensor motor commands so as to overcome usual limitations in the case of a static world : resolve ambiguities (e.g. front-back), recover observability of variables, etc. The study focuses on two-step strategies : (1) extraction of spatial information and source activity detection from a short-term analysis of audio streams ; (2) temporal assimilation of these data and fusion with the sensor motor commands in a stochastic filtering scheme.
Document type :
Theses
Complete list of metadata

https://tel.archives-ouvertes.fr/tel-01017770
Contributor : Arlette Evrard <>
Submitted on : Thursday, July 3, 2014 - 10:34:56 AM
Last modification on : Thursday, June 10, 2021 - 3:06:25 AM
Long-term archiving on: : Friday, October 16, 2015 - 4:57:19 PM

Identifiers

  • HAL Id : tel-01017770, version 1

Citation

Alban Portello. Localisation binaurale active de sources sonores en robotique humanoïde. Robotique [cs.RO]. Université Paul Sabatier - Toulouse III, 2013. Français. ⟨tel-01017770⟩

Share

Metrics

Record views

608

Files downloads

1208