The human-machine interaction (HMI) field is benefiting from the latest advances in wearable devices, sensing technologies, and artificial intelligence (AI) models. This is allowing the development of new applications in real-life, virtual and augmented environments, where AI is assuming a significant role especially considering strongly human-centered, real-time, and noisy scenarios. The main motivations and relevance of the Artificial Intelligence for Human-Machine Interaction (AIxHMI) workshop regard (i) focusing on human-centered approaches and perspectives, (ii) exploiting AI approaches to provide a better interaction between humans and machines, (iii) moving AI to pervasive technologies relying on wearable devices and portable technologies, and (iv) presenting novel AI methods and proposals that exploit heterogeneous data sources, describing the human environmental interaction in real, virtual, and augmented scenarios. Seven papers have been submitted to the second edition of AIxHMI. Out of these, six have been accepted for this volume as regular papers. Diverse fields of HMI were touched by these authors as well as by the three invited speakers.
mands to an external application. Let us suppose that one of these systems comprises the use of wearable and wireless devices and that a precise and immediate feedback should be provided to its user. Among the possible issues that could arise there are the presence of excessive noise, the need for fast and non-computationally demanding algorithms for real-time neural signal interpretation, and the necessity of ensuring the users’ safety as well as the protection of the collected data.
These challenges can be easily found in other control and sensing devices, where the correct integration of heterogeneous data may also be a concern, especially when facing multimodal sensing and when considering the environment a human user is immersed in. In general, a HMI system should be reliable, safe, eficient, and accepted by its users.
Concerning this last point, users’ emotional involvement is a key factor when interacting with a machine. Therefore, machines that are able to adapt to the emotional states of their users as well as to evoke specific emotions may provide better communication between the two actors of these systems.
Therefore, user experience has an important role in the development of HMI applications.
The Artificial Intelligence for Human-Machine Interaction (AIxHMI) 1 workshop wants to assemble contributors from universities, research institutes, and industries working in multidisciplinary fields that pertain but that are not limited to HMI, BCI, control systems, wearable sensing and devices, emotional intelligence, afective computing, human centered sensing and computing, human factors and ergonomics, user experience, interface and sensor design, virtual and augmented reality, and ethics and security in AI, having that the AI is a transversal discipline that influences all these aspects.
Seven submissions have been sent from Indian (1), Italian (3), Switzerland (1), and Norwegian (1) Institutions and Industries by 27 authors to the AIxHMI workshop and six have been accepted in this volume.
Accepted papers mainly pertained to three domains: electroencephalographic (EEG) signal processing and management, non-intrusive and multi-modal physiological signal sensing, and attention in virtual reality (VR).
In particular, Mauro Mezzini [
Artemisia Sarteschi and Domenico G. Sorrenti [
Instead, Gabardi et al. [5] proposed a novel frequency-based deep learning methodology to reduce noise in EEG signals presenting ocular or muscular artifacts, considering the noise spectral features to provide appropriate filters for separating these artifacts from the actual neural signal.
A diferent topic of HMI is explored by Facchin et al. [6], who exploit a VR environment to analyse the attention of participants while driving with diferent distractors. Diferent analyses have been performed to evaluate if brain stimulation can enhance attention in two groups of participants, i.e., elderly and young drivers.
Moving from attention to psychological stress detection, Marthinsen et al. [7] make a precise assessment of the optimal electrode subset to improve machine learning techniques both in terms of performances and computational costs. In particular, the authors collect EEG data from students before exam sessions and after holidays to have a clear overview of possible stressors.
Finally, Malaspina et al. [8] propose an experimental protocol to assess immersiveness while considering diferent user interfaces. Besides using self-assessment questionnaires, the authors propose the use of diferent behavioural and physiological signals to provide more honest indicators of game immersion by exploiting non-invasive and wearable devices.
Besides the oral presentation of the aforementioned papers, three invited speakers participated in the AIxHMI workshop: • Stefano Mazzoleni, Associate Professor at the Department of Electrical and Information
Engineering of the Polytechnic University of Bari (Italy). • Léa Pillette, Researcher at IRISA (France). • Francesco Ferrise, Full Professor at the Department of Mechanical Engineering of the
Politecnico di Milano (Italy).
Stefano Mazzoleni contributed with a speech entitled: "AI-driven robot-assisted neurorehabilitation: from experimental trials towards a safe, reliable and efective human-robot interaction". In particular, a focus on novel robotic systems for rehabilitation of upper and lower limbs presenting characteristics of motor relearning is provided, making a clear assessment on the eficacy of these systems. An important point is made on the necessity of having safe, reliable and efective treatments, while considering scientific, technological, ethical, legal and social challenges bounded to robot-assisted rehabilitation.
Another HMI system is presented by Léa Pillette, who focused on EEG-based Brain Computer Interfaces (BCIs) and highlighted the absence of reliable BCIs outside of laboratory environments. A possible mean to improve the reliability of these systems is identified in an eficient user training, during which the users learn the production of diferent brain patterns that can be recognised by a machine. This justifies the question entitling the delivered talk: "Towards assessing and improving brain-computer interface user-training?".
Notice that neurofeedback is considered a key element for correct user training and thus an investigation on the influence of neurofeedback on BCI usability is provided as an intersection between HMI and VR.
The speech of Francesco Ferrise entitled "Emotion Elicitation in Virtual Reality: Techniques and Applications" is focused on the last cited topic. The importance of emotions in everyday life is highlighted and the novel introduction of VR based emotion generation explained. In particular, the use of simulated environments is seen as a way to immerse a user, allowing to study the user’s emotional changes in diferent contexts, while being in a safe space. Francesco Ferrise gave us some tips and tricks to evoke emotions in VR and a look in future VR-based applications.
Acknowledgments The AIxHMI workshop co-chairs would like to thank all the Program Committee members for their reviewing and dissemination help: • Gloria Beraldo, Istituto di Scienze e Tecnologie della Cognizione - Consiglio Nazionale delle Ricerche (Italy). • Cesar Caiafa, Argentinean Radioastronomy Institute (IAR) - CONICET, University of
Buenos Aires (Argentina). • Giulia Cisotto, University of Milano-Bicocca (Italy). • Mirco Gallazzi, University of Insubria (Italy). • Ignazio Gallo, University of Insubria (Italy). • Shkurta Gashi, ETH AI Center (Switzerland). • Francesca Gasparini, University of Milano-Bicocca (Italy). • Alessandra Grossi, University of Milano-Bicocca (Italy). • Karmele Lopez de Ipiña, University of the Basque Country (Spain). • Pere Marti-Puig, Universitat de Vic - Universitat Central de Catalunya (Spain). • Marta Molinas, Norwegian University of Science and Technology (Norway). • Evangelos Niforatos, Delft University of Technology (The Netherlands). • Silvia Santini, Università della Svizzera-Italiana (Switzerland). • Agnese Sbrollini, Università Politecnica delle Marche (Italy). • Marta Maria Sosa Navarro, University of Milano-Bicocca (Italy).
• Sun Zhe, Juntendo University (Japan). [5] M. Gabardi, A. Saibene, F. Gasparini, D. Rizzo, F. A. Stella, A multi-artifact eeg denoising by frequency-based deep learning, in: Proceedings of the Second Workshop on Artificial Intelligence for Human-Machine Interaction (AIxHMI 2023) co-located with the 22th International Conference of the Italian Association for Artificial Intelligence (AI*IA 2023), CEUR Workshop Proceedings, CEUR-WS.org, 2023. [6] A. Facchin, S. La Rocca, V. Strina, L. Vacchi, C. Lega, S. Fontana, Assessing the impact of selective attention in a minimalist virtual reality driving simulator: An analysis of perceived experience and motion sickness, in: Proceedings of the Second Workshop on Artificial Intelligence for Human-Machine Interaction (AIxHMI 2023) co-located with the 22th International Conference of the Italian Association for Artificial Intelligence (AI*IA 2023), CEUR Workshop Proceedings, CEUR-WS.org, 2023. [7] A. J. Marthinsen, I. T. Galtung, A. Cheema, C. M. Sletten, I. M. Andreassen, Ø. Sletta, A. Soler, M. Molinas, Psychological stress detection with optimally selected eeg channel using machine learning techniques, in: Proceedings of the Second Workshop on Artificial Intelligence for Human-Machine Interaction (AIxHMI 2023) co-located with the 22th International Conference of the Italian Association for Artificial Intelligence (AI*IA 2023), CEUR Workshop Proceedings, CEUR-WS.org, 2023. [8] M. Malaspina, J. Amianto Barbato, M. Cremaschi, F. Gasparini, A. Grossi, A. Saibene, An experimental protocol to access immersiveness in video games, in: Proceedings of the Second Workshop on Artificial Intelligence for Human-Machine Interaction (AIxHMI 2023) co-located with the 22th International Conference of the Italian Association for Artificial Intelligence (AI*IA 2023), CEUR Workshop Proceedings, CEUR-WS.org, 2023.