The special track papers are presented during the main conference and will have the same review process (same submission and presentation guidelines).
- The submission deadline for special track papers is 15 March 2024 (please see Important Dates for details)
- Authors interested in participating in a special track session are kindly requested to select the relevant special track as one of the submission topics when submitting their papers.
This year, we have two Special Track sessions: (1) Artificial Identity and Emotion: Unravelling the Depths of Generative Affective AI and (2) HRD-BOND: The What, Why, How of Social Bonding in Human-Robot Dyads.
(1) Artificial Identity and Emotion: Unravelling the Depths of Generative Affective AI
Organizers
- Guy Laban, University of Cambridge, UK E-mail: gl538@cam.ac.uk
- Dimosthenis Kontogiorgos, Institute of Technology, U.S.A. E-mail: dkonto@mit.edu
- Minha Lee, Eindhoven University of Technology, the Netherlands E-mail: m.lee@tue.nl
Aim and Scope of the Special Session
This special session aims to critically explore the nuanced interplay of artificial emotion and identity within the advanced realms of affective computing and intelligent interactions. Diverging from foundational research in the field, this session specifically delves into how emerging AI technologies, especially Generative AI and Large Language Models (LLMs), are redefining the boundaries of artificial emotions and identities. Our focus is on the sophisticated and often overlooked aspects of this interplay, examining how these technologies not only mimic human emotional responses but also shape unique artificial personas. Therefore, we call research to submit groundbreaking research that explores the intricate dynamics between affective computing techniques and the capabilities of LLMs in creating sophisticated, emotionally intelligent interactions.
Our aim is to bring together interdisciplinary expertise from AI, human-computer interaction, psychology, neuroscience, and ethics. This session will explore not just how artificial agents interpret, adapt to, and represent complex emotional states but also delve into the mechanisms of emotional mimicry in AI, how human responses to these synthetic emotions, and how LLMs integrate with affective computing technologies to handle diverse tasks and contexts. We seek to highlight research that tests the boundaries of affective computing, considering both the technological innovations and the deep ethical and social ramifications of these advancements.
We invite submissions that showcase advanced research, innovative methodologies, and thought-provoking insights in the realm of affective computing and intelligent interactions, particularly focusing on the intricate relationship between artificial emotion and identity. Contributions should emphasize novel approaches, challenge existing paradigms, and provide deep insights into the complex dynamics of emotional AI. This session aims to foster a comprehensive and forward-thinking dialogue, setting a path for future ethical, social, and technical developments in this fascinating area of research.
List of Potential Topics
- Emotionally Nuanced Communication via LLMs in Interactions with Artificial Agents: Examining how Large Language Models enhance agents’ ability to convey emotionally rich and context-aware communications, influencing their perceived identity in human-agent interactions.
- Mechanisms of Emotional Mimicry in AI: Investigating the processes and technologies enabling AI systems to mimic complex human emotions accurately.
- Artificial Identity versus Human Identity: Exploring the nature of artificial identity in the context of LLM-enhanced robots and how it compares to human identity.
- Human-AI Emotional Interaction Dynamics: Exploring how humans perceive and interact with AI systems that exhibit emotionally intelligent behaviors and the psychological impact of these interactions.
- Synergy of LLMs and Affective Computing Techniques: Examining the integration of LLMs with traditional affective computing approaches and how this synergy enhances the emotional intelligence of AI systems.
- Contextual and Task-specific Emotional AI Responses: Investigating how AI systems adapt their emotional responses based on specific contexts or tasks and the effectiveness of these adaptations.
- Integration of Human Intelligence and Artificial Identity: Analysing how human intelligence is translated into artificial identity through code and hardware and how machines react to and synthesize human behavior.
- Emotional Expressivity of Artificial Agents: Investigating the delivery mechanisms of artificial identities through emotional expressions, including facial expressions, gestures, voice, speech, and content.
- Ethical and Social Implications of Emotionally Advanced AI: Delving deeper into the ethical considerations, including trust, empathy, and societal impact, arising from interactions with emotionally intelligent AI.
- Emotional AI in Diverse Applications: Exploring the use of emotionally intelligent AI across various sectors and settings, from healthcare to customer service, and its implications.
- Innovative Techniques in Emotion Synthesis and Recognition: Showcasing cutting-edge developments in AI systems’ synthesis and recognition of complex emotional states.
- Cross-Disciplinary Approaches to Emotional AI: Encouraging research that integrates insights from psychology, neuroscience, ethics, and philosophy with technological advancements in AI.
(2) HRD-BOND: The What, Why, How of Social Bonding in Human-Robot Dyads
Organizers
- Session Chair: Imran Khan (Imran.khan@ait.gu.se), University of Gothenburg (Sweden)
- Session Co-chair: Robert Lowe (robert.lowe@ait.gu.se)
Aim and Scope of the Special Session
We are heading towards a society where interactions between humans and social robots will no longer be reserved for rare occasions (e.g., in research facilities, exhibitions, or a one-shot experience in a public space). Rather, it will be a collaborative and assistive co-existence across many domains: including healthcare services, education, public spaces, and assisted transportation. The improvements in autonomy, intelligence, and behavioral capabilities of these social agents over time, combined with our ever-increasing (collaborative) interactions with them and our disposition as a social species, bring about the potential for humans to form strong, meaningful relationships with social robots.
In human-human and even human-animal dyads, social attachments, bonds, or relationships serve powerful regulatory roles on our emotions, behaviors, and mental and physical states. Forming social bonds is associated with increased levels of trust and perceived competence of our partners, increased levels of cooperation and collaboration between partners, and providing means for affective regulation (e.g., reducing anxiety). If these are qualities we are looking to imbue in social robots, and if they are to advance to the next stage as long-term collaborative, assistive social partners, then the phenomenology of our relationships with these artificial agents requires a deeper scholarly focus.
We believe the conference theme, “Human and Beyond”, necessitates this discussion. In this first edition of the HRD-BOND Special Session, we aim to provide a focused forum to collate work, ideas, and discourse on theoretical or applied work on this phenomenon. We welcome discussion on, for instance, understanding if, how, and why social bonds can form in human-robot dyads, establishing frameworks and methods to measure and evaluate their quality and functionality, and discourse on the ontology and epistemology of social interaction and bonding between humans and artificial systems, as well as ethical and legal considerations of these relationships with social robots. We also particularly encourage and welcome
transdisciplinary approaches (for example, work inspired by social psychology, ethology, neuroscience, evolutionary biology, anthropology, or other related disciplines), applied work related to social bonding in (long-term) human-robot studies, and position papers providing novel perspectives or challenges to existing ideas.
List of Potential Topics
- Applied work on social bonding or attachment in long-term human-robot (dyadic) interactions.
- (Adaptive) functionality of social bonding or attachments in human-robot scenarios, e.g., personalization of robot interactions.
- Behavioral consequences or expressions of social bonding/attachment dynamics in human-robot dyads (e.g., changes in spatial proximity, frequency of interactions, ‘personality’ changes of robot or human).
- Novel (theoretical or applied) frameworks, tools, or metrics to evaluating/measuring quantifying dynamics of social bonding or attachment in human-robot dyads.
- Discussions on ontology and epistemology of social bonding, attachment, or forming of relationships between human-robot dyads.
- Contextual (human) factors affecting social attachment dynamics or characteristics in human-robot interaction scenarios, e.g., cultural, social, and gender norms.
- Work discussing ethical, moral, and legal considerations and implications of social bonding or attachments with social robots or other (social) autonomous agents.