ABSTRACT. Single Image Super Resolution (SISR) is an active area of research within the deep learning community, aiming at enhancing the resolution of a low-resolution (LR) observation to produce its high-resolution (HR) counterparts.

We begin by examining various loss functions employed in SISR models, analyzing how they affect performance metrics commonly used to assess these methods. The evaluation of different metrics on the quality and accuracy of the generated HR images is critically evaluated using HR generated samples.

Then, we explore the significance of the training datasets, namely how the specificity of these datasets impacts the performance of SISR models. Both quantitative metrics and qualitative assessments of the HR samples are used to gauge the effectiveness of different training datasets.

Finally, the paper compares two distinct SISR approaches: the supervised approach, where models are pre-trained on extensive datasets before inference, and the unsupervised approach, which generates HR images from a single LR image without prior training.

ABSTRACT. It is widely recognised that reading is essential for children’s education. Nevertheless, the hours children should spend reading are increasingly being replaced by hours in front of tablets, phones, or computer screens. To counter this trend, we developed a tool that allows educators to create stories that combine children’s interests with literary classics. The goal is to motivate the reading indirectly through children’s interests. The tool exploits the generative capabilities of ChatGPT to create compelling crossovers between literature classics and children’s interests. The tool was developed under the design participatory paradigm. A user study conducted in Lisbon, Portugal with 20 children shows a positive experience with the tool and an increase in motivation to read the classics after using it. This paper presents the design, development, and evaluation of the tool.

ABSTRACT. The rapid evolution of generative artificial intelligence (GenAI) is revolutionising various areas, including education and gaming industry. GenAI can create original content to enhance traditional teaching methods, making learning more interactive and personalised. These tools can significantly improve educational outcomes by providing personalised feedback to students and increasing their engagement and motivation. However, the integration of GenAI in education raises ethical concerns, particularly regarding privacy, bias, and the accuracy of AI-generated content, as well as the authenticity and authorship of the work. There is a strong emphasis on the need for robust ethical guidelines and human oversight to mitigate these issues. In your study, we use GenAI to create an NPC with a unique personality and life background and enable learners to interact with the NPC without scripted dialogue, creating an engaging game-based learning environment. The prototype developed was evaluated by a group of sixteen students, and the main results are presented and discussed.

ABSTRACT. 3D point clouds, such as those produced by 3D scanners, often require labeling –- the accurate classification of each point into structural or semantic categories -- before they can be used in their intended application. However, in the absence of fully automated methods, such labeling must be performed manually, which can prove extremely time and labour intensive. To address this we present a virtual reality tool for accelerating and improving the manual labeling of very large 3D point clouds. The labeling tool provides a variety of 3D interactions for efficient viewing, selection and labeling of points using the controllers of consumer VR-kits. The main contribution of our work is a mixed CPU/GPU-based data structure that supports rendering, selection and labeling with immediate visual feedback at high frame rates necessary for a convenient VR experience. Our mixed CPU/GPU data structure supports fluid interaction with very large point clouds in VR, what is not possible with existing continuous level-of-detail rendering algorithms. We evaluate our method with 25 users on tasks involving point clouds of up to 50 million points and find convincing results that support the case for VR-based point cloud labeling.

ABSTRACT. This article presents a virtual fitting room system that improves customer experience by providing personalized image-based recommendations. The prototype designed takes a step forward in the state of the art by contemplating the creation of a customizable avatar based on the user's biometric characteristics and clothing style. Facial recognition models were implemented to predict gender and age, and segmentation and classification techniques are used to extract characteristics from the clothing the user is wearing. The work describes the progress and experiences in the development of some prototype modules and the possible methodologies that are being evaluated to develop a real-time web-based adaptation experience to represent the user's appearance and simulate a combination of multiple garments.

ABSTRACT. In the modern world, making healthy food choices is increasingly important due to the rise in food-related illnesses. Existing tools, such as Nutri-Score and comprehensive food labels, often pose challenges for many consumers. This paper proposes an application that uses image recognition technologies to read and interpret food labels, thus upgrading current solutions that rely mainly on reading product barcodes. By using advanced optical character recognition and machine learning techniques, the system aims to accurately extract and analyze nutritional information directly from food packaging without relying on a database of pre registered products. This innovative approach not only increases consumer awareness, but also supports personalized diet management for diseases such as diabetes and hypertension, and also promotes healthier eating habits and better health outcomes. Two minimalist functional prototypes were developed as a result of this work: a desktop application and a mobile application

ABSTRACT. This paper presents a technical proof-of-concept for utilizing cost-effective, computer vision-based pose estimation to enable natural and virtual navigation in VR-CAVEs. Utilizing MediaPipe for pose estimation we examined key challenges in applying pose data to navigation concepts. To address issues such as low transmission rate of pose data, joint position fluctuations, and difficulties in capturing fast movements, we implemented a processing pipeline that normalizes extremity lengths, filters pose data with a Kalman filter and classifies hand gestures using minimal training data. We developed a prototype (LCVR-CAVE) to test natural navigation through user-centred perspective adjustments based on filtered head positions, and virtual navigation concepts using gesture-based controls and 3D arm raycasts. The virtual navigation methods include teleportation, panning, flying, and view rotation, activated through specific hand gestures and movement patterns. An initial evaluation indicated that while natural navigation provided an immersive experience, the effectiveness of virtual navigation concepts varied depending on the implemented navigation approach, with low pose sampling rates having the most significant impact.

ABSTRACT. Prof. Sylvia Xueni Pan PhD is a full Professor of VR at Goldsmiths, University of London. She co-leads the Goldsmiths Social, Empathic, and Embodied VR lab (SeeVR Lab) and the MA/MSc in Virtual and Augmented Reality programme at Goldsmiths Computing. Her 2017 Coursera VR specialisation attracted over 100,000 learners globally. Her research interest is the use of Virtual Reality as a medium for real-time social interaction, in particular in the application areas of training and therapy.

Virtual Social Interaction

Virtual reality can give you the illusion of being with someone else, without being physically together. There are many practical examples of why this could be useful, and our capacity for building realistic virtual humans pushes the limits of new ways in which we could connect with each other, both in real life and virtually. I will discuss a few research projects covering a range of applications, including communication training, medical treatment, social neuroscience research, and immersive gaming. We will explore what social interaction means, and address the challenges posed by the rise of the AI-powered metaverse.

ABSTRACT. Virtual reality (VR) has the potential to significantly boost productivity in professional settings, especially those that can benefit from immersive environments that allow a better and more thorough way of visualizing information. However, the physical demands of mid-air movements make it difficult to use VR for extended periods. DeskVR offers a solution that allows users to engage in VR while seated at a desk, minimizing physical exhaustion. However, developing appropriate motion techniques for this context is challenging due to limited mobility and space constraints. This work focuses on object manipulation techniques, exploring touch-based and mid-air-based approaches to design a suitable solution for DeskVR, hypothesizing that touch-based object manipulation techniques could be as effective as mid-air object manipulation in a DeskVR scenario while less physically demanding. Thus, we propose Scaled Indirect Touch 6-DOF (SIT6), an indirect touch-based object manipulation technique incorporating scaled input mapping to address precision and out-of-reach manipulation issues. The implementation of our solution consists of a state machine with error-handling mechanisms and visual indicators to enhance interaction. User experiments were conducted to compare the SIT6 technique with a baseline mid-air approach, revealing comparable effectiveness while demanding less physical exertion. These results validated our hypothesis and established SIT6 as a viable option for object manipulation in DeskVR scenarios.

ABSTRACT. Technologies such as chroma-key, LED walls, motion capture (mocap), 3D visual storyboards, and simulcams are revolutionizing how films featuring visual effects are produced. Despite their popularity, these technologies have introduced new challenges for actors. An increased workload is faced when digital characters are animated via mocap, since actors are requested to use their imagination to envision what characters see and do on set. This work investigates how Mixed Reality (MR) technology can support actors during mocap sessions by presenting a collaborative MR system named CoMR-MoCap, which allows actors to rehearse scenes by overlaying digital contents onto the real set. Using a Video See-Through Head Mounted Display (VST-HMD), actors can see digital representations of performers in mocap suits and digital scene contents in real time. The system supports collaboration, enabling multiple actors to wear both mocap suits to animate digital characters and VST-HMDs to visualize the digital contents. A user study involving 24 participants compared CoMR-MoCap to the traditional method using physical props and visual cues. The results showed that CoMR-MoCap significantly improved actors' ability to position themselves and direct their gaze, and it offered advantages in terms of usability, spatial and social presence, embodiment, and perceived effectiveness over the traditional method.

ABSTRACT. The BraveHearts AR project aims to integrate Augmented Reality (AR) with Child Centered Play Therapy, a well-established strategy for mitigating emotions like anxiety, fear and stress, tailored to the specific needs and characteristics of children, as the target audience. Therefore, the development of a game using AR appeared as an attractive solution to explore, serving as a tool for pediatric patients to reduce negative emotions. More specifically, providing users with the positive experiences of relaxation, fun and companionship while educating them about the pre-surgical procedures that they will undergo. Aiming to develop the best solution possible, a related research about relevant subjects was conducted. A prototype of a game was developed using Unity software, designed for mobile devices and features an intuitive interface. With this, the game comprises an introduction, a memory game focused on the playful aspect, minigames that emphasize learning about three pre-surgical procedures, a mechanism planned to keep engaged when playing, and an ending. To enhance the clarity of the medical procedures and to prioritize a positive experience, a narrative was carefully written about the game, involving a buddy character that accompanies the player along the game. Throughout the game, the buddy introduces new characters who interact with the player while explaining medical procedures using simple game mechanics. The ambition of developing an AR game capable of transmitting information and educating its players about surgical procedures was fulfilled. Evaluations about the game’s usability were conducted and the great results obtained so far are motivating to further test this game with patients in appropriate places by the target audience, i.e., with children who are hospitalized for surgery, aiming at accessing its capability to reduce the aforementioned negative emotions.

ABSTRACT. This paper explores the integration of Augmented Reality (AR) and a 3D product viewer to enhance the furniture buying experience in an e-commerce setting. Utilizing ARCore and ARKit alongside Three.js, the project focused on usability and user experience (UX) through a thoughtful design approach. The prototype achieved positive usability testing outcomes by overcoming challenges such as mobile optimisation. This work contributes to advancing e-commerce for furniture stores through immersive technologies and emphasizes the importance of prioritizing UX. Future directions may involve further usability enhancements and scalability across devices.

ABSTRACT. Most collaborative immersive 3D content creation platforms do not consider the concept of accountability, which is a crucial aspect of collaborative tasks, providing a feeling of ownership and responsibility for the outcome of the creation. We propose an approach for 3D content creation versioning, enabling users to browse past changes and identify what operations were performed, who made them, and when. It aims to address the lack of accountability in immersive and collaborative content creation platforms. A user evaluation with 29 participants demonstrated that our approach does increase users' accountability within the virtual environment.

ABSTRACT. Augmented Reality applications enhance the physical world by overlaying digital components in an interactive 3D space. These applications use cameras and sensors to capture environmental data, which can raise privacy concerns for users and bystanders. This work focuses on the identification and protection of sensitive data in Location-Based Augmented Reality Games. We introduce the SafeARUnity library—a real-time image-processing middleware that sanitizes sensitive information before rendering. We address implementation challenges related to Unity Sentis and YOLO, with the aim of integrating SafeARUnity into Augmented Reality applications such as mobile games and other AR-enabled platforms. This is a work in progress, focused on testing and validating the effectiveness and reliability of the developed solution in preserving user privacy without compromising AR functionality.

ABSTRACT. This paper proposes a tool for automated conflict detection in smart home programming. The tool resorts to a model-checking approach to analyse trigger-action rules to detect potential problems.

ABSTRACT. Mixed reality simulators combine virtual and augmented reality that can create immersive training environments to enhance learning experiences and outcomes. Artificial Intelligence can play a transformative role in professional training, especially when integrated with Mixed Reality simulators. These advanced technologies combine to create immersive, interactive, and highly effective training environments across various industries.

ABSTRACT. Virtual reality (VR) technology has advanced significantly, enabling more immersive experiences by combining visual, auditory, and tactile inputs with walking platforms. This multisensory immersion stimulates multiple senses, enhancing the feeling of presence in the virtual world. Walking devices, like omni-directional treadmills, allow users to walk or run in VR while remaining stationary in real life. These devices are valuable for applications in training and rehabilitation, offering a more natural interaction compared to controllers. However, challenges such as high costs, space requirements, and motion sickness persist. Despite these issues, the potential of combining walking platforms with VR headsets is immense, particularly in fields like education and healthcare. As the technology becomes more accessible, it is expected to revolutionize multiple industries.

ABSTRACT. The alarming decline in populations of many shark species requires an increase in education on this critical environmental issue. Overexploitation and indiscriminate fishing have pushed many shark species to the brink of extinction, threatening marine ecosystem biodiversity and trophic structure. This paper describes the development of a serious game, comprising two mini-games, designed to positively influence children’s perceptions and raise awareness about the crucial roles sharks play in marine ecosystems. Additionally, the study aims to establish a link between player types and performance metrics collected during gameplay. The game was developed using a participatory design method, involving children throughout different phases of the project. It underwent four versions, with a total of 52 children participating. Out of these, 38 tested the final prototype, providing promising and interesting results both in the knowledge gain and in perception, which positively changes pos-game

ABSTRACT. Incidental visualizations allow individuals to access information on-the-go, at-a-glance, and without needing to consciously search for it. Unlike ambient visualizations, incidental visualizations are not fixed in a specific location and only appear briefly within a person’s field of view while they are engaged in a primary task. Despite their potential, incidental visualizations have not yet been thoroughly studied in current literature. We conducted exploratory research to establish the distinctiveness of incidental visualizations and to advocate for their study as an independent research topic. We tested both incidental and ambient visualizations in two scenarios: one involving a cognitively demanding primary task, and the other involving a mechanical primary task. Our findings show that in the cognitively demanding task, both types of visualizations resulted in similar performance. However, in the mechanical task, ambient visualizations led to better results compared to incidental visualizations. Based on these results, we argue that incidental visualizations are indeed distinct and merit further exploration, particularly in contexts where the primary tasks are more mechanically demanding.

ABSTRACT. Stroke survivors often face significant limitations in both mobility and cognitive function, causing the relearning of lost skills and abilities. To achieve the best possible quality of life, rehabilitation becomes an essential and ongoing process. However, the literature indicates that sustaining survivors' motivation and engagement is a substantial challenge when relying solely on traditional rehabilitation methods. These conventional approaches, often perceived as monotonous and repetitive, can lead to decreased patient adherence and suboptimal outcomes. Consequently, there is a pressing need for innovative rehabilitation strategies that can improve overall rehabilitation success. To help overcome this constraint, we explored a Virtual Reality (VR) supermarket in a Rehabilitation Center for five months. This prolonged study, usually not reported in literature, was done through a Human-Centered Design (HCD) methodology, in collaboration with healthcare professionals and stroke survivors. By simulating an environment where survivors can engage in tasks mirroring real-life activities, it was possible to foster cognitive and physical recovery. We report the impressions collected, suggesting that integrating VR into rehabilitation programs may be a valuable tool for addressing the multifaceted challenges faced by stroke survivors. We also detail the next steps of this research line, focusing on having an understanding of the long-term use of this VR-based approach in stroke rehabilitation.