Gestures and Postures in Social Signal Processing
Gestures and Postures in Social Signal Processing∗
Resul Collaku
Department of Informatics
Bulevard de Perolles 90
1700 Fribourg, Switzerland
resul.collaku@unifr.ch
ABSTRACT
This paper concentrates on two important types of behavioral cues, gestures and postures, how are they recognized, algorithms used for recognizing gestures and postures, in what kind of application areas are they used, their classification and interaction with Social Signal Processing (SSP).
In this context, a contribution is given to the approaches of researches that seek to find ways of gaining access to and devising automated systems for machine analysis of these social signals. To fulfill this need, this paper brakes down gestures and postures both descriptively and analytically and ultimately paves the way for future research.
1.
between the individuals, their stance, their role in the conversation (speaker/listener) as well as the level and type of commitment to the ongoing communication, etc. Additionally, we can go a step further and notice that there seem to be two social interactions, one between two people and one between three people, whereby each individual could be characterized by their gesticulation and their expression of distinctive gestures and postures.
INTRODUCTION
A central part of social intelligence depicted through social interactions plays the understanding of nonverbal interpersonal communication such as social signals and social behaviors [3][1]. In the aspect of computing, there arises the necessity to analyze and make use of social signals as part of human intelligence [2]. This becomes therefore the axle of research of Social Signal Processing as illustrated in the figure below.
Even though it might be easy for a human to infer the underlying story from the situation in figure 1, it becomes clear fast that these pieces of information are not really revealed from the figure itself, but instead are a result of our deduction and
Resul Collaku
Department of Informatics
Bulevard de Perolles 90
1700 Fribourg, Switzerland
resul.collaku@unifr.ch
ABSTRACT
This paper concentrates on two important types of behavioral cues, gestures and postures, how are they recognized, algorithms used for recognizing gestures and postures, in what kind of application areas are they used, their classification and interaction with Social Signal Processing (SSP).
In this context, a contribution is given to the approaches of researches that seek to find ways of gaining access to and devising automated systems for machine analysis of these social signals. To fulfill this need, this paper brakes down gestures and postures both descriptively and analytically and ultimately paves the way for future research.
1.
between the individuals, their stance, their role in the conversation (speaker/listener) as well as the level and type of commitment to the ongoing communication, etc. Additionally, we can go a step further and notice that there seem to be two social interactions, one between two people and one between three people, whereby each individual could be characterized by their gesticulation and their expression of distinctive gestures and postures.
INTRODUCTION
A central part of social intelligence depicted through social interactions plays the understanding of nonverbal interpersonal communication such as social signals and social behaviors [3][1]. In the aspect of computing, there arises the necessity to analyze and make use of social signals as part of human intelligence [2]. This becomes therefore the axle of research of Social Signal Processing as illustrated in the figure below.
Even though it might be easy for a human to infer the underlying story from the situation in figure 1, it becomes clear fast that these pieces of information are not really revealed from the figure itself, but instead are a result of our deduction and
References: (New York, NY, USA), pp. 185–218, 2008. through nonverbal behavior analysis,” in Computer Vision and Pattern Recognition Workshops, 2009. Serbia), 2011. Man, and Cybernetics (SMC), 2011 IEEE International Conference on, pp dimensions of physical space,” in Computer Vision and Pattern Recognition Workshops (CVPRW), 2010 IEEE Computer Society Conference on, pp. 51–58, june. 2010. devices and services on, pp. 513–516, 2010. and Analyzing Multimodality, Malta, May 2010 on, 2010. Conversations,” in Pattern Recognition (ICPR), 2010 20th International Conference on, pp