Part of the work done during the Research Master internship permit a publication at the international conference of
Content Based Multimedia Indexing (CBMI) in 2018 for the special session on
"Analysis of Multimedia Data for Medicine and Health".
This publication was presented at the conference at La Rochelle (FRA) on September 04, 2018
Co-authors: Pierre HANNA, Myriam DESAINTE-CATHERINE
Automatic Identification of audio caracteristics, application to pulmonary sounds
Research Master internship done in Studio for Creation and Research in Computer science and Experimental Music (SCRIME in french)
Full time internship from March, 15 2018 to July,31 2018
Internship responsible: Pierre HANNA, Co-responsible: Myriam DESAINTE-CATHERINE
The main goal of this internship was to try some machine learning mehods to classify respiratory sounds
depending of their audible symptoms. An approach with boosted decisional tree was tested.
A financial support was obtained to continu this work during a PhD. thesis.
Programming language: Python
Interface for piloting the "Thymio" robot with "I-score" software
Work done in Studio for Creation and Research in Computer science and Experimental Music (SCRIME in french)
Alternate internship from november,5 2015 to june,27 2016
Internship responsible: Pierre COCHARD, Co-responsible: Phillipe GUILLEM
The main goal of this internship was to create an interface permitting to pilot the "Thymio" robot with the i-score software to permit to use it in class with kindergarten children with Mr Phillipe GUILLEM.
To reach this goal, we implement un communication system between th Pyo library and i-score.
Programming language: C
The use of Haar cascades on 2D images for authentification on mobile devices.
Internship report:Work protected by a confidentiality agreement
Work done at the Laboratory of Bordeaux of Research in Computing Science (LaBRI) after a request from SOGETI.
Full-time internship from june,1 to july,29 2015.
Internship responsible: Pascal DESBARATS, Co-responsible: Maximilien OBERLIS
The goal was to work on an application for mobile devices already in development in the SOGETI society to make the authentification more reliable from a human face; using remarkables points.
Programming languages: Python, C, OpenCV library
