Human action recognition is concerned with analysing, understanding and synthesizing kinematic and kinetic data. Data-driven computer vision solutions have advanced the progress in the field of action recognition while it would not be possible without large and publicly available datasets. While there are numerous datasets available for adult action recognition, child-centered action recognition efforts are still scarce. This paper presents a dataset that consists of video recordings of 200 children of different age and gender groups performing seven actions (boxing, waving, clapping, running, jogging, walking towards the camera, and walking from side to side). Two camera perspectives are provided, with a top view and a frontal view in RGB format. The dataset provides an opportunity for kinematic analysis of children motion while it also aims to bridge the performance gap between activity recognition systems for adults and children.

Robot-Mediated Interventions (RMIs) promise to help autism specialists support children with Autism Spectrum Conditions (ASC). This practice focuses on developing robot-enhanced support systems in which social robots act as assistants and mediators in autism therapy. Understanding the perspectives and experiences of autism specialists is key to evaluating the added value of robots in educational and therapeutic settings. For this purpose, our team collaborated with an autism center, where over 100 children with ASC participated in RMI sessions one to two times per week for one year. As part of the study, we conducted in-depth interviews with two practitioners, exploring their attitudes toward embedding robots into their daily work and practices. We analyzed data on four key themes-acceptance of RMI, evaluations of RMI effects, procedural outcomes and potential improvements. The results highlight the positive acceptance of robots in autism therapy, although practitioners raised concerns about functional and technical limitations. The findings suggest practical considerations for researchers, practitioners, and robot developers in the design and implementation of RMIs.

Children bring unique perspectives and valuable input into design processes, unburdened by the complexities of societal influences. Their direct involvement in the design process is essential for creating meaningful, engaging, and inclusive activities. Participatory design (PD) research is becoming increasingly important in involving end-users in co-designing robotic systems and software. This paper presents the preliminary findings of a PD workshop aimed at establishing an interactive and collaborative environment for children’s co-design of a learning activity for robot-assisted learning on the Furhat robot. To this end, we conducted four workshop sessions with children aged 2-8 years old to co-create a robot-assisted learning scenario surrounding the topic of animals. As a result, together with the children, we designed and tested the prototype of a learning scenario, which helped us identify the shortcomings and set the foundation for future robot-assisted learning scenarios and activities. By reflecting on the challenges and lessons learned through PD with young children, we contribute to enhancing the understanding of PD.

Children's early literacy skills tend to unfold given the specially designed environment and learning activities. In light of Kazakhstan's ongoing alphabet shift from Cyrillic to Latin, many challenges arise for early literacy development and acquisition of the new script. Human-robot interaction (HRI) research provides a unique venue for the integration of social robots into language learning, while the Montessori method with a century-long hisory offers autonomous, selfdirected, and collaborative learning space. Combining these areas, we propose a robotic system named Moveable Älıpbi, designed to help boost children's motivation in learning the new script in long-term interaction. We detail the system's interaction design life cycle from understanding users to designing and implementing robot behaviors followed by evaluating with pilot participants.

Prior research on sign language recognition has already demonstrated encouraging outcomes in achieving highly accurate and dependable automatic sign language recognition. The use of virtual characters as virtual assistants has significantly increased in the past decade. However, the progress in sign language generation and output that closely resembles physiologically believable human motions is still in its early stages. This assertion explains the lack of progress in virtual intelligent signing generative systems. Aside from the development of signing systems, scholarly research have revealed a significant deficiency in evaluating sign language generation systems by those who are deaf and use sign language. This paper presents the findings of a user study conducted with deaf signers. The study is aimed at comparing a state-of-the-art sign language generation system with a skilled sign language interpreter. The study focused on testing established metrics to gain insights into usability of such metrics for deaf signers and how deaf signers perceive signing agents.

Social Narratives (SNs) have shown promising benefits for helping children navigate diverse social situations. The use of SNs in Robot-Assisted Play (RAP) can offer creative ways to address social and communication challenges of children with functional needs. In this study, we evaluate the initial use of social stories in robot-assisted therapy to help 17 children with diverse conditions such as Down Syndrome (DS), Autism Spectrum Conditions (ASC), and speech disorders (SD) understand and act on real-world situations while interacting with the social robot Furhat. We developed educational stories based on six common professions to show how each professional works and interacts with people. We conducted a user study as part of their daily intervention in a rehabilitation center over two weeks. Overall results demonstrate that there were no significant differences in socio-emotional outcomes among children grouped by their diagnosis, age, and other characteristics. However, their communication skills played a significant role in their performance; in particular, children with verbal abilities had higher completion time in activities compared to their non-verbal counterparts. This result was supported by two therapists who reported higher engagement with the robot among verbal children. We suggest that the SN-inspired intervention could be used to support children with diverse special needs, without causing any negative impacts on their learning.

The development of data-driven behaviour generating systems has recently become the focus of considerable attention in the fields of human–agent interaction and human–robot interaction. Although rule-based approaches were dominant for years, these proved inflexible and expensive to develop. The difficulty of developing production rules, as well as the need for manual configuration to generate artificial behaviours, places a limit on how complex and diverse rule-based behaviours can be. In contrast, actual human–human interaction data collected using tracking and recording devices makes humanlike multimodal co-speech behaviour generation possible using machine learning and specifically, in recent years, deep learning. This survey provides an overview of the state of the art of deep learning-based co-speech behaviour generation models and offers an outlook for future research in this area.

This research designed and applied 24 multi-purposeful robot activities of varying social mediation levels in a multiple-session experiment with 34 children of diverse autistic characteristics in a rehabilitation setting. This paper explores what type of robot activities can meet individual needs to bring more socio-behavioral progress and juxtaposes child characteristics to identify behavioral outcomes in each activity. This knowledge would help us to respond to the question of what activity types suit specific subgroups of Autism Spectrum Disorder (ASD). Our data analysis included coding 48.5 hours of video data for a total of 14 measures to fully capture children's activity-based socio-emotional outcomes. Overall, the activities on varying social mediation levels brought more or less positive social outcomes to all children. However, children showed some different behavioral outcomes as mediated by core autism-related and age-specific characteristics. This study provides in-depth accounts of what might be helpful in designing and applying multi-purposeful activities responsive to the diverse needs of children.

The present study applies a novel Reinforcement Learning-based (RL) alphabet learning system named QWriter for the acquisition of the Kazakh Latin alphabet. We conducted a between-subject design experiment with 108 Kazakh children aged 6-8 years old in a public school and compared their learning rates across the two conditions: an RL-based QWriter robot and a human tutor (HT) as a baseline. The results show that children learned significantly more letters with the HT compared to the QWriter robot, showing that the RL-based robot is not effective for learning in the short term. Yet, we observe some interesting results by children’s age and gender. The results need further investigation comparing the QWriter with other robot baselines with different roles and across various learning tasks.

This paper studies the perception of robot intelligence for a manipulator that shares its workspace with a human operator, while the two execute independent tasks. Four different motion planning algorithms were employed to plan the robot motion in four subsequent experimental sets for 48 participants, and the order with which these algorithms were used was changed, obtaining all possible combinations. While guaranteeing safety, each of the four algorithms exhibited a different level of adaptability and reactivity, in terms of real-time motion planning abilities and of speed reduction based on heart rate feedback. We analyzed how perceived intelligence was influenced by the employed algorithm, by the order of execution and by the previous experience of the participants working with robots. In conclusion, perceived intelligence resulted being significantly lower in the first experimental set compared to the following ones, which showed a positive effect of habituation on perceived intelligence. On the other hand, neither the use of different motion planning algorithms nor the previous participants’ experience significantly influenced perceived intelligence.

Sign Languages (SL) are a form of communication in the visual-gestural modality, and are full-fledged natural languages. Recent years have witnessed the increase in the use of virtual avatars as virtual assistants. Research into sign language recognition has demonstrated promising potential for robust automatic sign language recognition. However, the area of sign language synthesis is still in its infancy. This explains the underdevelopment of virtual intelligent signing systems. Additionally, existing models are often restricted to manually written rules and require expert knowledge, while data-driven approach could provide a better solution. Apart from the development of signing systems, research indicates a gap in the evaluation thereof by sign language users. In this paper, we propose a data-driven sign language interpreting avatar and its subjective evaluation. We present findings from a pilot study with the deaf evaluating two different avatars against a human sign language interpreter using the metrics that are believed to bring out important insights and narratives for the users in terms of their perceptions of the avatars.

The application of Reinforcement Learning (RL) as an emergent field of Machine Learning has shown positive results in interdisciplinary fields. Although research has proven its effectiveness in language education through various agents (e.g., chatbots, robots, talking avatars), its application in letter acquisition is relatively new. In light of the alphabet transition from Cyrillic to Latin for the Kazakh language, potential challenges might be associated with learning and memorizing the new alphabet. Specifically, students with extant alphabet knowledge might struggle in using the later-learnt new alphabet given no sufficient practice. In this paper, we present a chatbot based on Reinforcement Learning that is anticipated to assist university students in learning the Kazakh Latin alphabet during an interaction in a letter acquisition scenario. Thus, we attempt to identify whether the RL chatbot is efficient for this learning scenario through an online survey study involving pre-test, chatbot interaction, and post-test.

Social robots are increasingly being used as a medi-ator in robot-assisted autism therapy to improve children's social and cognitive skills. Engagement is one of the key measurements used to evaluate the therapeutic interventions' effect on children. While “engagement” is broadly used, it has been challenging to find a consensus about its definition in the community. With this paper, we explore the use of a data-driven approach to investigate the extent to which a model on engagement built on one dataset transfers to another. We utilized two publicly available datasets of engagement recognition, namely PInSoRo and Qamqor datasets, with an attempt to achieve a higher accuracy taking into account the transferred knowledge. The accuracy of 83.18% was obtained on the PInSoRo dataset of child-child interactions with face and body keypoints. We have used the methodology of transfer learning to improve the classification accuracy on the Qamqor dataset. The best result obtained is a 71.89% accuracy on the Qamqor dataset. This suggests that more data with similar keypoints is needed to achieve better accuracy when utilizing transfer learning from one dataset to another dataset.

Research has recognized the importance of individ-ual differences of children with Autism Spectrum Disorder (ASD) that require interventions to meet their heterogeneous needs. This relatively large-scale study investigates a robot-assisted autism therapy (RAAT) with 34 children with diverse forms of ASD and Attention Deficit Hyperactivity Disorder (ADHD). We conducted a multi-session study with multi-purposeful activities targeting the socio-emotional abilities of children in a rehabilitation setting. We found a number of quantitative results suggesting various autism-related and demographic differences such as diverse forms of ASD, co-occurrence of ADHD, verbal skills, and age groups. The main findings are: 1) severity of ASD forms may not predict intervention outcomes but instead the co-occurrence of ADHD with LFA diagnosis may negatively impact social smiling; 2) verbal children were more generally engaged and less aggressive with the robot than non-verbal children whose curiosity rose over sessions; and 3) younger children (3.4 y.o.) showed more affection, while older children (7–12 y.o.) were better engaged through speaking more words and having longer engagement and eye contact with the robot.

This paper presents the methodology we used to crowdsource a data collection of a new large-scale signer independent dataset for Kazakh-Russian Sign Language (KRSL) created for Sign Language Processing. By involving the Deaf community throughout the research process, we firstly designed a research protocol and then performed an efficient crowdsourcing campaign that resulted in a new FluentSigners-50 dataset. The FluentSigners-50 dataset consists of 173 sentences performed by 50 KRSL signers for 43,250 video samples. Dataset contributors recorded videos in real-life settings on various backgrounds using various devices such as smartphones and web cameras. Therefore, each dataset contribution has a varying distance to the camera, camera angles and aspect ratio, video quality, and frame rates. Additionally, the proposed dataset contains a high degree of linguistic and inter-signer variability and thus is a better training set for recognizing a real-life signed speech. FluentSigners-50 is publicly available at https://krslproject.github.io/fluentsigners-50/

This paper presents a new handwritten dataset, Cyrillic-MNIST, a Cyrillic version of the MNIST dataset, comprising of 121,234 samples of 42 Cyrillic letters. The performance of Cyrillic-MNIST is evaluated using standard deep learning approaches and is compared to the Extended MNIST (EMNIST) dataset. The dataset is available at https://github.com/bolattleubayev/cmnist

Social robots deployed in the therapy of autism is a promising and important research domain. Recently, an increasing amount of work is being conducted utilizing a social robot as a mediator between a therapist and a child with autism. Being able to evaluate how engaged a child is both offline and in real-time would improve the quality of the provided robot-assisted intervention and also provide objective metrics for later analysis by the therapist. The state-of-the-art engagement recognition is challenged by the diverse styles of expressing engagement by this vulnerable population group. To this end, this PhD project aims to explore how transfer learning can improve the recognition accuracy of children's engagement with the robot or another human. We will utilize four publicly available multi-modal datasets to discover a suitable feature representation of engagement during various types of activities with the robot.

Research on child-agent interaction is rapidly expanding. It is, therefore, necessary to converge our collective efforts to broaden our understanding and perspectives of how virtual agents, affect and potentially improve the well-being of children. “Children, Robots and Virtual Agents: Present and Future Challenges” follows our International Conference on Social Robotics (ICSR) 2020 workshop on child-robot interactions. In this full-day workshop, we will focus on the unique technical and empirical challenges of designing and conducting child-agent interactions. In light of the current pandemic situation, we will also address the challenges and adaptations of conducting research under the “new normal” to understand how researchers overcome these challenges and what we can learn and keep in the future. We also aim to join the virtual agents and robotics communities to learn from each other and discuss both areas’ common and specific challenges. Our primary goal is to provide an opportunity for an interdisciplinary debate about the present and future of child-agent interactions. We want to bring together researchers, practitioners and pioneers from relevant disciplines and create collaboration opportunities. As part of the workshop, we will have a collaborative activity where our participants will work together and brainstorm about intelligent agents in different time frames (past, present and future). We will also have a panel of experts discussing the topics of this workshop and answering participants questions.

3D avatars are widely used in various fields of emerging technology, from augmented reality to social robots. To interact in a natural way with the user, they must be able to show at least some basic emotions. However, generating animation for these virtual avatars is a time-consuming task and a creative process. The main goal of this work is to facilitate a generation of facial animation of basic emotions on a 3D avatar. To this end, we developed and compared two approaches. The first method consists of the generation of animation using tuning Blendshape features of the 3D model, whereas the second method captures it from the real face and maps it on the model correspondingly. Additionally, the text, from which the emotion was estimated, was passed to lip synchronization software for generating realistic lip movements for the avatar. Then, animations of six basic emotions were shown in different variations in the survey and respondents were asked to guess the emotion shown in the video. Besides, such anthropomorphic features of the avatar as human-likeness, life-likeness and pleasantness were examined. In general, the analysis of the survey provided the following interesting findings: a) participants did not have significant differences in recognizing emotions based on the type of animation generation method; b) inclusion of voice significantly enhanced the recognition of emotion. In relation to participants’ accuracy of emotion recognition, Excitement and Happiness were mostly confused between each other more than any other two emotions, while Anger was the easiest emotion to recognize.

The paper presents the first dataset that aims to serve interdisciplinary purposes for the utility of computer vision community and sign language linguistics.To date, a majority of Sign Language Recognition (SLR) approaches focus on recognising sign language as a manual gesture recognition problem.However, signers use other articulators: facial expressions, head and body position and movement to convey linguistic information.Given the important role of non-manual markers, this paper proposes a dataset and presents a use case to stress the importance of including non-manual features to improve the recognition accuracy of signs.To the best of our knowledge no prior publicly available dataset exists that explicitly focuses on non-manual components responsible for the grammar of sign languages.To this end, the proposed dataset contains 28250 videos of signs of high resolution and quality, with annotation of manual and nonmanual components.We conducted a series of evaluations in order to investigate whether non-manual components would improve signs' recognition accuracy.We release the dataset to encourage SLR researchers and help advance current progress in this area toward realtime sign language interpretation.

In the Republic of Kazakhstan, the transition from Cyrillic to Latin alphabet raises challenges to training an entire population in writing the new script. This paper presents a CoWriting Kazakh system, an extension of the existing CoWriter system, aiming to implement an autonomous social robot that would assist children in transition from the old Cyrillic alphabet to a new Latin alphabet. With the aim to investigate which learning strategy yields better learning gains, we conducted an experiment with 67 children, aged 8-11 years old, who interacted with a robot in a CoWriting Kazakh learning scenario. Participants were asked to teach a humanoid NAO robot how to write Kazakh words using one of the scripts, Latin or Cyrillic. We hypothesized that a scenario in which the child is asked to mentally convert the word to Latin would be more effective than having the robot perform conversion itself. Results show that the CoWriter was successfully applied to this new script-switching task. The findings also suggest interesting gender differences in the preferred method of learning with the robot.

This paper aims to develop a deep neural network for Continuous Sign Language Recognition (CSLR) with iterative Gloss Recognition (GR) fine-tuning. CSLR has been a popular research field in the last few years and iterative optimization methods are well established. This paper introduces our proposed architecture involving Spatiotemporal feature-extraction model to segment useful “gloss-unit” features and BiLSTM with CTC as a sequence model. Spatiotemporal Feature Extractor is used for both image features extraction and sequence length reduction. To this end, we compare different architectures for feature extraction and sequence model. In addition, we iteratively fine-tune feature extractor on gloss-unit video segments with alignments from the end2end model. During the iterative training, we use novel alignment correction technique, which is based on minimum transformations of Levenshtein distance. All the experiments are conducted on the RWTH-PHOENIX-Weather-2014 dataset.

The motivation behind this work lies in the need to differentiate between similar signs that differ in non-manual components present in any sign. To this end, we recorded full sentences signed by five native signers and extracted 5200 isolated sign samples of twenty frequently used signs in Kazakh-Russian Sign Language (K-RSL), which have similar manual components but differ in non-manual components (i.e. facial expressions, eyebrow height, mouth, and head orientation). We conducted a series of evaluations in order to investigate whether non-manual components would improve sign’s recognition accuracy. Among standard machine learning approaches, Logistic Regression produced the best results, 78.2% of accuracy for dataset with 20 signs and 77.9% of accuracy for dataset with 2 classes (statement vs question). Dataset can be downloaded from the following website: https://krslproject.github.io/krsl20/

The Robots for Learning workshop series aims at advancing the research topics related to the use of social robots in educational contexts. This year's half-day workshop follows on previous events in Human-Robot Interaction conferences focusing on efforts to design, develop and test new robotics systems that help learners. This 5th edition of the workshop will be dealing in particular on the potential use of robots for adaptive learning. Since the past few years, inclusive education have been a key policy in a number of countries, aiming to provide equal changes and common ground to all. In this workshop, we aim to discuss strategies to design robotics system able to adapt to the learners' abilities, to provide assistance and to demonstrate long-term learning effects.

The gradual transition towards the Kazakh language in the Republic of Kazakhstan raises the emergence of applying new technologies for learning the language. Considering the fact that the Kazakh language has dialectal forms, it is important to investigate how these language features would affect the interaction with the synthesized speech of a robot or a computer program. This paper presents a preliminary study exploring the effect of dialectal language on the human-robot interaction in an education-oriented environment. Participants were involved in the interaction with two different robots with pre-programmed language dialectal patterns - South and non-South, to learn new vocabulary. Findings show that there is a low significance in correlation, however, it is suggested that a small sample size led to the obtained results.

In the Republic of Kazakhstan, the transition from Cyrillic towards Latin alphabet raises challenges to teach the whole population in writing the new script. This paper presents a CoWriting Kazakh system that aims to implement an autonomous behavior of a social robot that would assist children in learning a new script. Considering the fact that the current generation of primary school children have to be fluent in both Kazakh scripts, this exploratory study aims to investigate which learning approach provides better effect. Participants were asked to teach a humanoid robot NAO how to write Kazakh words using one of the scripts, Latin vs Cyrillic. We hypothesize that it is more effective when a child mentally converts the word to Latin in comparison to having the robot perform conversion itself. The findings reject this hypothesis, but further research is needed as it is suggested that the way the pre-test was performed might have caused the obtained results.

This paper presents a study that aims to investigate the effects of Robot-Assisted Therapy (RAT) on children who have a form of verbal development retardation such as Delayed Speech Development (DSD). To this end, we developed a number of applications for a humanoid robot NAO with the aim to engage children during RAT sessions. We conducted an evaluation of these applications with DSD children who interacted with the robot on a few occasions. Our findings demonstrate the utility of such applications for the therapy of DSD children which was both engaging and entertaining. Similar approach could be utilized for the therapy of children with Autism Spectrum Disorder and Attention Deficit Hyperactivity Disorder.

Deaf-mute communities around the world experience a need in effective human-robot interaction system that would act as an interpreter in public places such as banks, hospitals, or police stations. The focus of this work is to address the challenges presented to hearing-impaired people by developing an interpreting robotic system required for effective communication in public places. To this end, we utilize a previously developed neural network-based learning architecture to recognize Cyrillic manual alphabet, which is used for fingerspelling in Kazakhstan. In order to train and test the performance of the recognition system, we collected four datasets comprising of static and motion RGB and RGB-D data of 33 manual gestures. After applying them to standard machine learning algorithms as well as to our previously developed learning-based method, we achieved an average accuracy of 93% for a complete alphabet recognition by modeling motion depth data.

Robots deployed in public spaces must necessarily deal with situations that demand them to engage humans in a socially and culturally appropriate manner. However, social environments are often complex and ambiguous: many queries to the robot are collaborative (e.g. a family), and in case of conflicting queries, social robots need to participate in value decisions and negotiating multi-party interactions. Given the strong influence of the people's demographic information and social schema among people, such as relationships and hierarchies, the focus of this research is to examine whether and how people exhibit socio-psychological effects with a shared robot deployed at international events or spaces (e.g. airports). With the aim to investigate who robots should adapt to (children or adults) in multi-party situations within human-robot interactions in public spaces and whether this adaptation can be influenced by culture, this paper presents a cross-cultural study conducted online. The results include a number of interesting findings based on people's relationship with a child and their parental status. In addition, a number of cross-cultural differences were identified in respondents' attitude towards robot's multi-party adaptation in various public settings.

This research explores the concept of edutainment where the basics of programming are introduced to children while playing a game with a social humanoid robot. The goal of the game is to exit the maze: the child is asked by the robot to make the robot walk the maze to its exit. The child needs to learn the basics of programming to walk through the maze via drag-and-drop instructions on the tablet screen. This paper presents an HRI study which aims to investigate what role of the robot (peer vs. teacher) would result in more learning gains in this particular application. The findings suggest that children complete the task much quicker with the peer robot while a teacher robot is shown to be more effective for learning.

This paper presents a study that aims to address the challenges of creating effective educational robots. To this end, we developed a social educational robot which acts as a peer that is also a learner of a foreign language rather than a tutor teaching it. Children engage in a game with a peer robot which is programmed to either always win or always lose. The purpose of this study is to investigate whether children would learn more if they were winning or losing the game to the robot. The conducted study compares children's responses and children's improvements of English vocabulary after a game with the robot. Results indicate that children improve their English after playing with the robot, however which depends on the robot's adapting strategy.

Robots in public environments are challenged with socially appropriate interactions with previously unseen users: they need to offer appropriate services and shape their interaction style according to the particular individual's needs and preferences, for example the elderly and children. In addition, interactions in public spaces are not limited to merely two parties, but often involve multi-party situations with changing numbers of participants. The research question of this work is to investigate what rules should a socially competent robot follow in order to adapt to such complex social situations in real-world scenarios.

This paper presents an observational study conducted at the public environment investigating children's engagement with a social robot within a pretend play scenario. A pretend play system was designed in order to address the challenges of evaluating child-robot interaction by exploiting the advantages of ubiquitous robotics and ambient intelligence of multimodal, multi-sensory perception. The system performed successfully at a children's play centre where a humanoid NAO robot was able to dynamically adapt its gender by changing its gendered voice to match the gender of the child. By analyzing the free play of children, the results obtained confirm the hypothesis of gender segregation within child-robot interaction. These findings are important and ought to be considered when designing robotic applications for children in order to increase robot's perceived likeability, acceptance and engagement.

This paper presents a study that aims to address the challenges of creating effective educational robots. To this end, we developed a social educational robot which acts as a peer that is also a learner of a foreign language rather than a tutor teaching it. Children engage in a game with a peer robot which is programmed to either always win or always lose. The purpose of this study is to investigate whether children would learn more if they were winning or losing the game to the robot. The study was conducted with 76 children aged 6-10 years old. The conducted study compares children's responses and children's improvements of English vocabulary after a game with the robot. Results indicate that children improve their English significantly more when playing with the losing robot. In addition, girls rated playing with the robot significantly higher than boys did regardless of the robot's playing strategy.

In order to inform the design of robotic applications for children, in this paper we describe and report the results of an experiment we conducted in a primary school. Our work investigates the effects of the robot's perceived gender and age on levels of engagement and acceptance of the robot by children across different age and gender groups. Our results show that children across ages relate differently toward perceived robot's age and gender.

Service robots employed in public spaces need to be equipped with specific sensing, reasoning and human-robot interaction capabilities to adapt their interaction style and thus effectively engage with a variety of users. In this paper we present a method used by an ubiquitous robotic system to gather 3D body metrics and use them to robustly estimate age and gender of previously unseen participants in real-world multi-party situations. We evaluate system's performance on 428 children volunteers and compare them with those obtainable with a state of the art software based on face analysis. This work demonstrates that even small number of biometrics can achieve good age and gender estimation results in perceptually challenging environments. Finally, this paper illustrates how the system is used to inform the online adaptation of the behavior of a humanoid robot.

Networked robotic applications enable robots to operate in distant, hazardous, or otherwise inaccessible environments, such as search and rescue, surveillance, and exploration applications.