Master 2018 2019
Stages de la spécialité SAR
Socially Intelligent Computing for Groups’ Analysis


Site : Socially Intelligent Computing for Groups’ Analysis
Lieu : Laboratoire Traitement et Communication de l’Information”, Telecom ParisTech, Paris, France
Encadrant : Giovanna Varni (giovanna.varni@telecom-paristech.fr)
Dates :du 1/03/2019 au 1/09/2019
Rémunération :full-time intern fellowship according to the Telecom ParisTech salary scale
Mots-clés : Master SAR, autre qu’ATIAM, Parcours ATIAM : Traitement du signal

Description

An intern position extendable in a 3-years fully funded PhD student position is now available at “Laboratoire Traitement et Communication de l’Information”, Telecom ParisTech, Paris, France (https://www.telecom-paristech.fr/en...)

Topic ******* Socially Intelligent Computing for Groups’ Analysis

Supervisor ********* Prof Giovanna Varni

Keywords ******** Social Signal Processing, Affective Computing, Information Technology, multimodal data analysis, multimodal dataset, machine learning, deep learning, computer vision, experimental design.

Framework : ********** The internship (and the possible thesis) will focus on the development of a computational model of cohesion in groups’ of humans, able to integrate task and social dimensions and also accounting for their development over time. This work will be conducted in the framework of the 3-years long ANR JCJC French national project GRACE (Groups’ Analysis for automated Cohesion Estimation) starting in mid-late 2019.

Recently, with the advent of Social Signal Processing aiming at developing socially intelligent machines [1], Computer Science research took interest in investigating and quantifying relational phenomena involving two or more persons. Groups of people are a fascinating interdisciplinary phenomenon. They can be defined as bounded and structured entities that emerge from the purposive, interdependent actions of individuals.

One of the current open challenges on automated groups’ analysis is to provide computational models of higher level concepts called “emergent states”, that is, states emerging as the result of affective, behavioral and cognitive interaction among the members of a group [2]. Cohesion is one of these states considered as a highly valued group property serving crucial roles for group effectiveness and performance.

Scholars proposed theoretical models of cohesion having from one to five dimensions [3]. Among these dimensions, the task and social ones were the most investigated. The task dimension concerns the extent to which group members are united to achieve the group’s goals and objectives ; the social dimension refers to the social relationships within the group (e.g. the extent to which group members like each other.

Tasks ***** The intern will work on several tasks :

1) Studying the state-of-the-art on cohesion to identify which are its most suitable and frequent multimodal behavioral descriptors. State-of-the-art will span several research fields, including sociology, psychology, and computer science

2) Computing multimodal behavioral descriptors of cohesion

4) Designing, implementing, and evaluating a computational model of cohesion

Candidate profile ************** The ideal candidate should have a strong background on Computer Science, AI, Information Technology, Applied Mathematics or closely related fields is preferable. In addition to a passion for science and programming, the candidate should be open to approach and solve the issues linked to Human-Computer Interaction.

The following skills are also expected :
- Good command of English (written and spoken). French language is not a mandatory requirement
- Interest in multidisciplinary research at the interface between computer science and sociology/psychology
- Proof of Excellent student career
- Strong programming skills (e.g. C++/Python)
- Very good communication skills, commitment, independent working style as well as initiative and team spirit

Offer : ***** Starting date : March/Avril 2019.

Salary : full-time intern fellowship according to the Telecom ParisTech salary scale.

Application deadline : the evaluation of the candidates starts immediately and it will continue until the position is filled.

Application ********* To apply please send by email to giovanna.varni@telecom-paristech.fr in a single pdf file :
- A cover letter stating your research interests and how they could be related to the research topic the thesis focuses on.
- A detailed CV
- Transcripts of student records of your MSc

For any additional questions about the position, please contact Prof. Giovanna Varni (giovanna.varni@telecom-paristech.fr). Please quote “Intern position” in the email subject for both asking information and application.

Bibliographie

[1] Vinciarelli, A., Pantic, M., and Bourlard, H. 2009. Social signal processing, Image and Vision Computing, 27, 12, pp. 1743-1759. [2] Nanninga, M. C., Zhang, Y., Lehmann-Willenbrock, N., Szlavik, Z., and Hung, H. 2017. Estimating verbal expressions of task and social cohesion in meetings by quantifying paralinguistic mimicry, in Proceedings of the 19th ACM International Conference on Multimodal Interaction, pp. 206-215. [3] Salas, E., Grossman, R., Hughes, A. M., and Coultas, C. W., 2015. Measuring team cohesion : Observations from the science. Human Factors, 57, 3, pp. 365-374.