An increasing number of human-robot interaction (HRI) studies are now taking place in applied settings with children. These interactions often hinge on verbal to effectively achieve their goals. Great advances have been made adult speech recognition and it is assumed that these will carry over the HRI domain In this paper, we evaluate a automatic (ASR) engines under variety conditions, inspired by real-world social conditions. Using data collected demonstrate there still much work be done...

Companion robots are aimed to mitigate loneliness and social isolation among older adults by providing emotional support in their everyday lives. However, adults’ expectations of conversational companionship might substantially differ from what current technologies can achieve, as well other age groups like young adults. Thus, it is crucial involve the development companion ensure that these devices align with unique experiences. The recent advancement foundation models, such large language...

The field of Human-Robot Interaction (HRI) lies at the intersection several disciplines, and is rightfully perceived as a prime interface between engineering social sciences. In particular, our entertains close ties with cognitive psychology, there are many HRI studies which build upon commonly accepted results from psychology to explore novel relation humans machines. Key this endeavour trust we, field, put in methodologies it exactly that now being questioned across and, by extension,...

? To answer this question we designed and conducted a real-world long-term study, in collaboration with medical specialists, at the Fundación Cardioinfantil-Instituto de Cardiología clinic (Bogotá, Colombia) lasting 2.5 years. The study took place within context of outpatient phase patients' cardiac rehabilitation programme aimed to compare progress adherence conventional (

<title>Abstract</title> This work aims to provide initial guidelines towards developing companion robots with large language models (LLMs) be part of everyday lives older adults. Using iterative participatory design (co-design) approaches, we analyze the challenges applying LLMs for multi-modal open-domain dialogue, deriving from adults' (one-to-one) interactions a personalized robot, built on Furhat robot GPT-3.5. An study 6 Swedish-speaking adults (65 and older) showed that frequently...

For practical reasons, most human-robot interaction (HRI) studies focus on short-term interactions between humans and robots. However, such do not capture the difficulty of sustaining engagement quality across long-term interactions. Many real-world robot applications will require repeated relationship-building over long term, personalization adaptation to users be necessary maintain user build rapport trust robot. This full-day workshop brings together perspectives from a variety research...

Cardiovascular disease is the leading cause of death in world. A program cardiac rehabilitation (CR) related to physical activities or exercises regain optimal quality life. CR relies on necessity evaluate, control and supervise a patient's status progress. This work has two objectives: one hand, provide tool for clinicians assess during CR. On other there evidence that robots can motivate patients therapeutic procedures. Our sensor interface explores possibility integrate robotic agent into...

Turn-taking is a fundamental aspect of conversation, but current Human-Robot Interaction (HRI) systems often rely on simplistic, silence-based models, leading to unnatural pauses and interruptions. This paper investigates, for the first time, application general turn-taking specifically TurnGPT Voice Activity Projection (VAP), improve conversational dynamics in HRI. These models are trained human-human dialogue data using self-supervised learning objectives, without requiring domain-specific...

While most of the research in Human-Robot Interaction (HRI) focuses on short-term interactions, long-term interactions require bolder developments and a substantial amount resources, especially if robots are deployed wild. Robots need to incrementally learn new concepts or abilities lifelong fashion adapt their behaviors within situations personalize with users maintain interest engagement. The "Lifelong Learning Personalization Long-Term (LEAP-HRI)" Workshop aims take leap from traditional...

Conversation is one of the primary methods interaction between humans and robots. It provides a natural way communication with robot, thereby reducing obstacles that can be faced through other interfaces (e.g., text or touch) may cause difficulties to certain populations, such as elderly those disabilities, promoting inclusivity in Human-Robot Interaction (HRI). Work HRI has contributed significantly design, understanding evaluation human-robot conversational interactions. Concurrently,...

Real-world studies allow for testing the limits of HRI systems and observing how people react to failures. We developed a fully autonomous personalised barista robot deployed on an international student campus five days. experienced several challenges, most important one being speech recognition failures due foreign accents. Nonetheless, these showed different perspective HRI, we demonstrate personalisation can overcome negative user experience.

This paper presents a longitudinal case study of Robot Assisted Therapy for cardiac rehabilitation. The patient, who is 60-year old male that suffered myocardial infarction and received angioplasty surgery, successfully recovered after 35 sessions rehabilitation with social robot, lasting 18 weeks. took place directly at the clinic relied on an exercise regime which was designed by clinicians delivered support robot sensor suite. monitored patient's progress, provided personalised...

While earlier research in human-robot interaction pre-dominantly uses rule-based architectures for natural language interaction, these approaches are not flexible enough long-term interactions the real world due to large variation user utterances. In contrast, data-driven map input agent output directly, hence, provide more flexibility with variations without requiring any set of rules. However, generally applied single dialogue exchanges a and do build up memory over conversation different...

User identification is an essential step in creating a personalised long-term interaction with robots. This requires learning the users continuously and incrementally, possibly starting from state without any known user. In this article, we describe multi-modal incremental Bayesian network online learning, which first method that can be applied such scenarios. Face recognition used as primary biometric, it combined ancillary information, gender, age, height, time of to improve recognition....

The complex and largely unstructured nature of real-world situations makes it challenging for conventional closed-world robot learning solutions to adapt such interaction dynamics. These challenges become particularly pronounced in long-term interactions where robots need go beyond their past continuously evolve with changing environment settings personalize towards individual user behaviors. In contrast, open-world embraces the complexity unpredictability real world, enabling be "lifelong...

Social robots are demonstrating to have potential in several healthcare applications, especially rehabilitation areas. This paper presents an architecture for a socially assistive robot system cardiac rehabilitation, based on model-controller structure through finite-state machine and behaviour module. The platform has been designed provide social support assistance during the therapy, aiming improve quality of provided service, as well engagement performance patients. tested under clinical...

In order to achieve more believable interactions with artificial agents, there is a need produce dialogue that not only relevant, but also emotionally appropriate and consistent. This paper presents comprehensive system models the emotional state of users an agent dynamically adapt utterance selection. A Partially Observable Markov Decision Process (POMDP) online solver used model user reactions in real-time. The decides content next based on rewards from agent. previous approaches are...

Socially Assistive Robots (SAR) have been gaining significant attention in multiple health care applications. However, SAR has not fully explored cardiac rehabilitation (CR). One of the most critical issues CR is lack adherence patients to process. Hence, based on evidence that presence an embodied agent increases compliance, we present this paper integration a social robot programme. The setup evaluated with four divided into two conditions (robot and no robot), order evaluate its first...