| IEEE HOME | IEEE WEB ACCOUNTS | IEEE XPLORE | SMC SOCIETY | CONTACT IEEE |
IEEE Office 3 Park Avenue, 17th Floor New York, New York 10016-5997 U.S.A. Tel: +1 212 419 7900 Fax: +1 212 752 4929
Technical Committee on Social Signal Processing
Social Signal Processing (SSP) is the domain aimed at modelling, analysis and synthesis of nonverbal behaviour in social interactions (both human-human and human-machine). The key idea of the field is that non-verbal behavioural cues (facial expressions, vocalisations, postures, gestures, etc.) are the physical, machine detectable evidence of social and psychological phenomena not otherwise observable. Therefore, automatic detection and interpretation of nonverbal cues make it possible for machines to understand social phenomena (e.g., personality, conflict, interpersonal attraction, etc.). Furthermore, the synthesis of nonverbal behavioural cues via different forms of embodiment (robots, artificial agents, synthetic voices, etc.) makes it possible for machines to achieve the same social effects as humans, including persuasion, politeness, humor, etc.
The Technical Committee on Social Signal Processing accompanies the development of the field since its earliest steps. The TC chairs have been organizing the very first events dedicated to SSP in the scientific community, including the Social Signal Processing Workshops (2009-2011), the Workshop on Foundations of Social Signals (2009), the International Workshop on Socially Intelligent Surveillance and Monitoring (2010-2013), and the IEEE International Conference on Social Computing (2010-2012). Such an extensive activity has gathered a community of researchers around the topic that, until now, were scattered across different fields.
In parallel, the TC has worked to lay the theoretic foundations of the domain. After the seminal paper where the very expression "Social Signal Processing" was coined , the TC chairs refined the research vision of the domain through several, widely accepted position papers . Furthermore, TC members established solid interdisciplinary collaborations with human sciences by co-editing special issues centred around SSP in cognitive sciences ("Special Issue on Social Signals. From theory to application", Cognitive Processing) as well as multimodal interaction journals ("Special Issue on Conceptual frameworks for Multimodal Social Signal Processing", Journal on Multimodal User Interfaces).
The impact of such a foundational work was reinforced through the synergy with the SSPNet, the Europeam Network of Excellence on Social Signal Processing coordinated by the TC chairs. The SSPNet portal (https://www.sspnet.eu), featured in the "Best of Web" column of the IEEE Signal Processing Magazine , was instrumental to disseminate not only papers and initiatives mentioned above, but also resources crucial to attract new researchers to SSP, namely annotated data (roughly 250 hours), software packages (around 25) and roughly 150 presentation recordings collected at the various events organised, among others, by the TC members.
Furthermore, the collaboration with the SSPNet has allowed the TC members to co-organise the first international benchmarking campaigns on SSP related tasks, including the "Facial Expression Recognition and Analysis Challenge" (2011-2012), the "Speaker Trait Challenge" (2012), and the "Computational Paralinguistics Challenge" (2013). This has not only attracted a large number of researchers to SSP (more than 100 participants in total), but also diffused SSP-specific experimental practices and evaluation criteria, an important step towards the maturity of the domain.
After the foundational work above, the TC aims now at forming the new generation of SSP researchers through two initiatives specifically targeted towards PhD students and postdocs at the earliest stages of their career. The first initiative is the "Summer School on SSP" that will be held in Vietri Sul Mare (Italy) in June . The school will gather around 50 participants that will attend eight short courses about the most important aspects of SSP, from face and language cognition to social robotics and computational paralinguistics. The lectures will be recorded and made available online via the Virtual Learning Centre of the SSPNet (https://sspnet.eu/virtual-learning-centre-2/). In this way, the courses will be available not only to the school participants, but also to the entire scientific community. The second initiative is a textbook edited by the TC chairs in collaboration with Nadia Thalmann and Judee Burgoon. The book will be a collection of surveys that will cover not only the main areas of SSP (conceptual modelling, automatic analysis and synthesis of social signals), but also the most important applications of the domain so far. The book, expected for the end of 2014, will be the first volume including all aspects of SSP, a multifaceted and interdisciplinary domain that typically requires resources from multiple fields.
Finally, TC chairs and members contribute to move the large body of knowledge accumulated in the last years towards application in real-world problems. This is the goal, e.g., of the DCAPS project (led by one of the TC chairs) that aims at developing general metrics of psychological health . Besides demonstrating the societal impact of Social Signal Processing, DCAPS involves several universities and further expands the TC-SSP community.