ASIMOV 2024

ASIMOV - Adaptive Social Interaction based on user’s Mental mOdels and behaVior in HRI -
will be held on the 23th of October in conjunction with the 16th International Conference on Social Robotics 

23-26 October 2024, Odense (DK).

abstract

The ability to understand and adapt to people’s mental models is a key objective for enabling natural, efficient, and successful human-robot interaction (HRI), in particular in human-centered scenarios where robots are expected to meet people’s social conventions. Theory of mind and mental models are largely investigated in human-computer interactions, however, it is still unclear what level of others’ mental states a robot should be aware of in order to communicate with people in a transparent and socially acceptable way. The ASIMOV workshop will constitute a unique opportunity to gather roboticists and computer scientists to discuss a variety of current and new approaches aiming at endowing social robots with learning abilities, enhancing cognitive and social abilities based on mutual understanding (i.e. shared Theory of Mind).

SCIENTIFIC CONTEXT

Despite the promise of social robots in various domains, users often remain cautious about employing them due to ethical, psychological, and safety concerns. Addressing these acceptability challenges requires considering users' psychological and behavioral in the design of social interaction. Therefore, endowing robots with learning and online adaptation abilities is essential for improving human-robot interaction (HRI), especially in assistive, rehabilitation, and educational contexts.  Understanding and adapting to users' mental states can help address mismatches between expectations from the robots and their actual capabilities, thereby enhancing interaction efficiency. Such mismatches can lead to ambiguous perceptions and misinterpretations of robot actions, negatively impacting interactions. Recent research shows that robots' acceptability increases when they can understand and meet people's expectations during HRI. By equipping robots with basic socio-cognitive skills, they can convey contextually appropriate affective and social signals in an intelligent and readable way From the mutual comprehension of mental states, an effective HRI can emerge, suspending the disbelief of human partners, and allowing trust, partnership, and acceptability.  

OBJECTIVES

This workshop aims to unite theories and practices that enhance social cognition and user awareness in HRI, particularly for socially assistive robots in education, entertainment, and healthcare, where user acceptability is critical. Topics will be explored from a multidisciplinary perspective, inviting experts in human-agent interaction, social robotics, cognitive sciences, artificial intelligence, psychology, neuroscience, and philosophy. Special focus will be given to the state-of-the-art user modeling methods assessing overt (e.g., behavior and speech) and covert information (e.g., cognitive states and emotional reactions) using tools like motion capture, eye-tracking, and biosignals.

Target Audience

We aim to bring together a diverse audience from the fields of social and assistive robotics, cognitive and behavioral robotics, and HRI. The workshop will serve as a platform for exchanging ideas, discussing innovative concepts, and addressing unresolved issues in ongoing research. We encourage the participation of PhD students and young researchers working on user modeling, HRI and control interfaces, machine learning, and ethical aspects of human-machine interaction, among others. 

LIST OF TOPICS

 Topics of interest include, but are not limited to, the following:

• Mental models in HRI

• Human-aware perception-action loop

• Emotion and intention recognition

• Affective and cognitive sciences for socially interactive robots 

• Empathy and Theory of Mind in Robotics

• Mutual affective understanding

• Real-time monitoring of behavior and mental states

• Detection of non-verbal behavioral cues

• Online adaptive behavior

• Multimodality in human-robot interaction

• Acceptability and personalization

• Physiological monitoring and biofeedback systems

• BCI (brain-computer interfaces)-enabled adaptive interaction

• Short- and Long-term personalization

• Human partnership and trust in HRI

• Explainable AI in HRI

• Security and safety in HRI

news                                                                                                                                                                                                                                 

• Prof. Raphaëlle Roy, Université de Toulouse, France, has been confirmed as an invited speaker 

• Prof. Brian Scassellati, Yale University, USA, has been confirmed as an invited speaker