As virtual reality (VR) simulators become increasingly central to technical training, realistic haptic feedback is often assumed to enhance user experience, skill acquisition, and transfer to real-world tasks. This study investigates the impact of haptic feedback on basic technical motor skill training in VR, focusing on early-stage skills involving bimanual coordination, fine motor control, and precise tool-mediated object placement under visual and temporal constraints, with limited force modulation requirements. An immersive simulator implementing a bimanual “Ring Transfer” task was developed, requiring participants to precisely grasp and transfer small objects using two tools under visual guidance and timing constraints, with minimal force modulation. A closely matched physical setup replicating workspace dimensions, tool geometry, and interaction mechanics was used to assess skill transfer. Twenty-four novice participants were assigned to three groups: haptic (HC), no-haptic (NC), and control (CC), consistent with an a priori power analysis. The HC and NC groups completed nine training sessions over two weeks, followed by post-test and delayed retention evaluations conducted one week later. Performance was measured using task completion time and error rate, and user experience was assessed using the System Usability Scale (SUS) and a validated Presence Questionnaire. Results show that the NC group initially outperformed the HC group, likely due to increased cognitive and sensorimotor demands associated with adapting to force feedback. However, the latter demonstrated greater performance improvement over time. Both groups reported high usability, while the HC group reported significantly higher scores on the haptics dimension of the Presence Questionnaire. In the physical post-test, both trained groups significantly outperformed the control group, confirming effective transfer from VR to real-world performance. However, no significant difference was found between the HC and NC groups, indicating that the increased perceived realism associated with haptic feedback did not translate into measurable performance gains for this task. These findings highlight a task-dependent role of haptic feedback. In this study, basic tasks are characterized by a predominant reliance on visual guidance and spatial precision, with limited requirements for force control or tactile discrimination. For such tasks, visual information alone may be sufficient to support skill transfer. This suggests that the benefits of haptics in VR training depend on task demands, particularly the extent to which force control and tactile cues are required, whereas tasks involving significant force modulation or interaction with deformable objects may benefit more from haptic feedback.

In the early stages of learning a technical skill, trainees require guidance from a mentor through augmented feedback to develop higher expertise. However, the impact of such feedback and the different modalities used to communicate it remain underexplored in immersive virtual environments (IVE). This paper presents a study in which 27 participants were divided into three groups to learn a tool manipulation trajectory in an IVE. Two experimental groups received guidance from an expert using visual and/or haptic augmented feedback, while the control group received no feedback. The results indicate that both experimental groups showed significantly greater improvement in tool trajectory performance than the control group from pre- to post-test, with no significant differences between them. Analysis of their learning curves revealed similar performance improvements in tool trajectory across trials, outperforming the control group. Additionally, the visual-haptic feedback condition was linked to lower task load in three out of six dimensions of the NASA-TLX and a higher perceived interdependence with the expert's actions. These findings suggest that augmented feedback from an expert enhances the learning of tool manipulation skills. Although adding haptic feedback did not lead to better learning outcomes compared to visual feedback alone, it did enhance the overall user experience. These results offer valuable insights for designing IVEs that support mentor-trainee interactions through augmented feedback.

With the increasing use of immersive simulators in surgical training, there is an emerging emphasis on providing realistic haptic feedback. Although haptic feedback is recognized as necessary for surgical skill acquisition, its role in transferring those skills to real-life situations still needs to be explored. Our study aims to investigate the impact of haptic feedback on basic microsurgery skill acquisition, transfer, and retention, as well as on the user experience during VR training. An immersive simulator was developed to practice a microgrid task under magnification with and without haptic feedback. A similar physical setup with a binocular microscope was also designed to measure skills transfer to the real world. Thirty-three volunteers ($N=33$) were randomly divided into three participant groups: the haptic feedback group (HG), the no-haptic feedback group (NG), and the control group (CG). All participants performed the task on the physical setup during the pre-post-retention tests. After the pre-test, the first two groups performed six training trials on the immersive simulator, with the HG group receiving haptic feedback while performing the task. The control group did not receive any training. The results show that the HG and NG groups significantly improved their learning curve during training, with no significant differences between them. On the Other hand, the haptic feedback led to significantly higher usability and possibility of examination scores than those without haptic feedback. Finally, all the groups improved their time performance on the physical simulator. In addition, the haptic group participants showed a more significant gain in performance in terms of accuracy and error rates. These findings confirm the effectiveness of immersive environments combined with haptic feedback as a valuable training tool, facilitating the transfer of technical skills to realworld applications. Moreover, the results indicate that haptic feedback can also improve the user experience in immersive surgical simulators.

We introduce a compact video-classification pipeline for real-time dynamic hand-gesture recognition in mixed-reality (MR) settings. The network marries a MobileNetV3 backbone with two purpose-built temporal components: (1) a Gated Discriminative Temporal Shift Module (G-DiTSM) that inserts first-order motion differences and learns channel-wise gates to fuse them adaptively, and (2) a lightweight Depth-Efficient Channel Attention (DepthECA) block that recalibrates spatial features on the fly. Operating on eight sparsely sampled frames per clip (Temporal Segment Network paradigm), the resulting model contains 2.65 M parameters and requires only 0.084 GFLOPs per inference. Evaluated on the RGB-only 20BN Jester benchmark (148k clips spanning 27 gesture classes) recorded from front-facing viewpoints. The system reaches 95.34% Top-1 and 99.80% Top-5 accuracy, surpassing recent 3D CNNs and transformer baselines while being an order of magnitude lighter. Ablations confirm that DepthECA and G-DiTSM provide complementary gains (+18.78% and +0.93% Top-1, respectively, over the MobileNetV3 baseline). Because all components are plug-and-play and introduce minimal overhead, the architecture is well suited to the tight latency and power budgets of standalone MR headsets, paving the way for natural grab, rotate, and command interactions using only on-board RGB cameras

This study investigates the impact of two immersive debriefing strategies on learning outcomes in a virtual reality (VR) simulation focused on mobile cybersecurity. The simulation highlights everyday mistakes to raise awareness of risky behaviors in public settings. We developed and evaluated an immersive debriefing system that enables participants to review their performance and re-engage with the scenario through a redo phase. This redo was implemented in two formats: a full scenario redo (F-REDO) and a segmented redo targeting specific moments (S-REDO). We found that both redo formats were equally effective in learning outcomes and user satisfaction, using a mixed-methods approach combining standardized questionnaires (motivation, cognitive load, usability, knowledge retention) and qualitative trainer feedback. However, S-REDO demonstrated greater time efficiency without increasing cognitive load and was perceived by the trainer as more engaging and pedagogically relevant. These results support the integration of personalized, interactive debriefing tools in VR learning environments, particularly in domains requiring targeted remediation and critical decision-making. Abstract Abstract

We propose a novel object detection framework tailored for mixed reality (MR), combining a customizable synthetic dataset with a lightweight attention-enhanced detection model. Our dataset generation pipeline synthesizes planetary and telescope foregrounds with hybrid real-synthetic backgrounds, enabling robust learning across variable lighting and occlusion scenarios—challenges common in educational MR environments. At the core of our architecture is the Dual Focus Multiscale Attention (DFMA) module, which simultaneously refines spatial and channel-wise features at multiple scales. Integrated into a YOLO-based (You Only Look Once) backbone and FPN, DFMA significantly improves feature discrimination while preserving real-time efficiency. On MS COCO our model improves mean Average Precision (mAP) across Intersection over Union (IoU) thresholds from 0.5 to 0.95 (mAP@0.5:0.95) over state-of-the-art nano detectors from 39.3% to 41.3% (± 2%) at only +6% params and +3% GFLOPs, with a notable reduction in false positives on visually similar, low-textured objects. We further demonstrate real-time deployment in a Unity-based MR application, highlighting the system's effectiveness in immersive astronomy-focused educational scenarios. Our results underscore the potential of synthetic data and multiscale attention to bridge accuracy, speed, and realism in next generation MR systems.

Introduction Alzheimer's disease (AD) is the first cause of dementia worldwide without any current curative treatment. Facing an increasing prevalence and its associated costs, AD represents a public health challenge. Usual diagnostic methods still rely on extended interviews and paper tests provided by an exterior examiner. We aim to create a novel, quick cognitive-screening tool on a numerical tablet. Methods This pilot program, built and edited with Unity ® , runs on Android ® for the Samsung Galaxy Tab S7 FE ® . Composed of seven tasks inspired by the Mini-Mental Status Examination and the Montréal Cognitive Assessment, it browses several cognitive functions. The architectural design of this tablet application is distinguished by its multifaceted capabilities, encompassing not only seamless offline functionality but also a mechanism to ensure the singularity of data amalgamated from diverse sites. Additionally, a paramount emphasis is placed on safeguarding the confidentiality of patient information in the healthcare domain. Furthermore, the application empowers individual site managers to access and peruse specific datasets, enhancing their operational efficacy and decision-making processes. We performed a preliminary usability assessment among young, healthy subjects. Results Twenty-four participants were included with a final F-SUS ‘excellent’ score. Participants perceived the tool as simple to use and achieved the test in a mean time of 142 s. No technical errors occurred. Conclusion These preliminary results suggest that our new assessment on a numerical tablet might be usable and acceptable for a short cognitive screening but requires further studies among older populations.

In VR training, debriefing is as crucial as the simulation phase. With the growing adoption of these pedagogical tools, defining optimal educational approaches to maximize benefits for trainers and learners becomes essential. However, despite their benefits, VR-adapted debriefing methods still need to be explored. This paper presents an adaptable, all-in-one immersive debriefing module. It includes a complete system for recording, reviewing, and redoing actions. An ongoing study explores the redo module's effect in dynamic training scenarios. This module could enhance learning and reinforce immersive debriefing systems' interest and relevance.

Introduction: Alzheimer's disease (AD) is the first cause of dementia worldwide without any current curative treatment. Facing an increasing prevalence and its associated costs, AD represents a public health challenge. Usual diagnostic methods still rely on extended interviews and paper tests provided by an exterior examiner. We aim to create a novel, quick cognitive-screening tool on a numerical tablet. Methods: This program, built and edited with Unity®, runs on Android® for the Samsung Galaxy Tab S7 FE®. Composed of seven tasks inspired by the Mini-Mental Status Examination and the Montréal Cognitive Assessment, it browses several cognitive functions. The architectural design of this tablet application is distinguished by its multifaceted capabilities, encompassing not only seamless offline functionality but also a mechanism to ensure the singularity of data amalgamated from diverse sites. Additionally, a paramount emphasis is placed on safeguarding the confidentiality of patient information in the healthcare domain. Furthermore, the application empowers individual site managers to access and peruse specific datasets, enhancing their operational efficacy and decision-making processes. We performed a preliminary usability assessment among healthy subjects. Results: 24 healthy patients were included with a final F-SUS score of "excellent". Participants perceived the tool as simple to use and achieved the test in a mean time of 142 seconds. No technical errors occurred. Conclusion: These preliminary results suggest that our new assessment on a numerical tablet allows a short cognitive screening.

Virtual reality shows great promise as a technology for training healthcare professionals within a secure simulated environment. This work presents the design, development, and assessment of UltRASim: an immersive simulator for ultrasound-guided regional anesthesia. First, task and skills analyses were performed with domain experts to build the task model of the procedure and determine the simulator’s learning objectives and design constraints. Then, a face and content validity study was conducted with eighteen anesthesiologists to assess the simulator’s prototype. The responses to seven of eleven face validity questions were predominantly positive, indicating a favorable reception. The primary concerns pertained to the fidelity of haptic feedback during needle insertion. This suggests incorporating a higher fidelity haptic device in future design iterations. Conversely, responses to all six questions related to the content validity were predominantly positive. Participants found that the simulator held significant potential as a training tool, particularly for developing hand–eye coordination skills. These findings validate several design choices and highlight areas for improvement in subsequent iterations of UltRASim before its formal validation as a training tool.

We study the impact of haptic feedback on basic technical skills transfer from VR to the real world. Twenty-four volunteers were divided into two training groups (haptic and no-haptic groups) and a control group. The training groups learned to perform a “Ring Transfer” task in a VR simulator, and all participants performed pre-, post, and retention tests on a similar physical setup. Results show that skill transfer is observed for both training groups and not for the control group. The haptic group participants also improved their performance compared to the no-haptic group, but the difference was not significant.

With the mentoring model, a mentee can learn technical skills under the supervision of more experienced peers who demonstrate their knowledge through several communication modalities. Supporting the mentoring model within shared immersive training simulators holds promise in enhancing mentor-mentee interactions and learning outcomes in a safe environment. However, efficient communication within these spaces remains an open issue. This work presents a user study that explores the combination of communication modalities (verbal-visual, verbal-haptic, visual-haptic, and verbal-visual-haptic) to convey instructions to learners on the amplitude of movements to perform during a tool-handling task in an immersive environment. The study aims to examine the impact of the four modality combinations on performance (speed and accuracy of movement replication), mental workload, and participants’ user experience. The results show that participants achieved higher accuracy with the visual-haptic and verbal-visual-haptic conditions. Moreover, they performed the movements faster, and their movement trajectories were closer to the reference trajectories in the visual-haptic condition. Finally, the most preferred verbal-visual-haptic combination enhanced the users’ sense of presence, co-presence, social presence, and learning experience. No impact on the mental workload was observed. These results suggest that combining haptic and visual modalities is the best suited for enhancing learners’ performance. Adding the verbal modality can also improve the user experience in the immersive learning environment. These findings contribute to improving the design of immersive collaborative systems and pave the way for exploring novel avenues of research into the efficacy of multimodal communication for enhancing the mentoring-based acquisition of technical skills in VR. These tools hold promise for diverse applications, including medical simulation.

Simulation and debriefing are two essential and inseparable phases of virtual reality training. With the widespread adoption of these training tools, it is crucial to define the best pedagogical approaches for trainers and learners to maximize their effectiveness. However, despite their educational benefits, virtual reality-specific debriefing methods remain underexplored in research. This article proposes an architecture and interface for an all-in-one immersive debriefing module that is adaptable to different types of training, including a complete system for recording, replaying, and redoing actions. A study with 36 participants compared this immersive debriefing system with traditional discussion-based and video-supported debriefing. Participants were divided into three groups to evaluate the effectiveness of each method. The results showed no significant differences between these debriefing methods across several criteria, such as satisfaction, motivation, or information retention. Immersive debriefing is as usable and retentive as traditional or video debriefing in this context. The next step will be to evaluate the redo system in other training courses involving more dynamic scenarios.

La formation par compagnonnage permet aux novices d'acquérir des compétences sous la supervision d'experts qui utilisent diverses modalités de communication. Cependant, reproduire ce modèle dans des simulateurs immersifs reste un défi, notamment pour assurer une communication efficace entre experts et novices. Notre étude explore l'impact de la communication multimodale expert-novice pour transmettre des instructions sur l'amplitude des mouvements dans une tâche de manipulation d'outils en environnement immersif. Les résultats révèlent que la combinaison des modalités visuelle-haptique améliore la précision, la vitesse et la qualité des mouvements. De plus, la combinaison verbale-visuelle-haptique renforce le sentiment de présence et de coprésence, et l'expérience d'apprentissage. Ces résultats suggèrent que la combinaison visuelle-haptique est optimale pour améliorer les performances des novices, et que l'intégration de la modalité verbale améliore l'expérience utilisateur. Ces conclusions ouvrent de nouvelles perspectives pour améliorer l'acquisition de gestes techniques par compagnonnage en réalité virtuelle grâce à la communication multimodale.

Alzheimer's disease (AD) is the first cause of dementia worldwide and represents a public health challenge. Current diagnostic methods still rely on extended interviews and paper tests. We aim to create a novel, quick cognitive-screening tool on a numerical tablet. This program, built and edited with Unity®, runs on Android® for the Samsung Galaxy Tab S7 FE®. Composed of seven tasks inspired by the Mini-Mental Status Examination and the Montréal Cognitive Assessment, it browses several cognitive functions. The architectural design of this tablet application is distinguished by its multifaceted capabilities, encompassing not only seamless offline functionality but also a mechanism to ensure the singularity of data amalgamated from diverse sites. Additionally, a paramount emphasis is placed on safeguarding the confidentiality of patient information in the healthcare domain. Furthermore, the application empowers individual site managers by allowing them to access and peruse specific datasets, thereby enhancing their operational efficacy and decision-making processes. We performed a usability assessment among 24 healthy patients with a final F-SUS score of "excellent". Participants perceived the tool as simple to use and achieved the test in a mean time of 142 seconds, confirming that a short assessment on a numerical tablet is possible.

Real-time object detection in indoor settings is a challenging area of computer vision, faced with unique obstacles such as variable lighting and complex backgrounds. This field holds significant potential to revolutionize applications like augmented and mixed realities by enabling more seamless interactions between digital content and the physical world. However, the scarcity of research specifically fitted to the intricacies of indoor environments has highlighted a clear gap in the literature. To address this, our study delves into the evaluation of existing datasets and computational models, leading to the creation of a refined dataset. This new dataset is derived from OpenImages v7[14], focusing exclusively on 32 indoor categories selected for their relevance to real-world applications. Alongside this, we present an adaptation of a CNN detection model, incorporating an attention mechanism to enhance the model's ability to discern and prioritize critical features within cluttered indoor scenes. Our findings demonstrate that this approach is not just competitive to existing state-of-the-art models in accuracy and speed but also opens new avenues for research and application in the field of real-time indoor object detection.

Virtual and augmented reality technologies offer opportunities to create training and rehabilitation tools in virtual or augmented environments, giving users a first-person experience. Designing such systems requires defining the most critical components to be simulated, and to what degree they can be successfully simulated, considering the context of their use and the effectiveness of interactions. There are several research questions to which we have provided answers, including: How can the user be represented in the virtual environment, and what is the added value of this representation for medical training purposes? How can interaction fidelity impact user performance in surgical simulation-based training? What is the impact of augmented reality feedback on the performance of a patient's gait rehabilitation process? In addition, I will present two rehabilitation assistance systems for improving patient usability and motivation: 1) for upper limb motor rehabilitation after stroke in virtual reality; 2) gait rehabilitation in augmented reality.

This paper presents a specialized Single Shot Multi-box Detector (SSD) [9] tailored to the unique challenges of detecting low-textured objects in complex scenes. We harness the Fruit 360 dataset [10], featuring low-textured images of various fruits and vegetables, for both training and validation purposes. Our primary objective is to enable the integration of this streamlined SSD model [9] into mobile devices for the enhancement of mixed and augmented reality experiences. Remarkably, our custom model delivers a fourfold increase in processing speed compared to the original SSD [9], all while preserving or even enhancing detection accuracy on the designated dataset. Implementation is carried out in Python using TensorFlow

This work investigates the potential benefits of using a shared immersive environment for training purposes. Such an environment provides a safe space for teachers to impart their knowledge and expertise to trainees, especially when teaching technical skills that require proper tool manipulation. Our research focuses on exploring different communication modalities that can be used to teach movement amplitude during tool manipulation tasks. Specifically, we examine the effectiveness of haptic, visual, and verbal modalities in enhancing the learning process. Our user study results reveal that trainees were able to replicate movements more accurately when given instructions using the visual modality, and they were able to replicate movements faster when given instructions using the haptic modality. While verbal instructions increased the sense of copresence with the teacher, it was the least preferred modality. These findings suggest that multimodality could be the most appropriate approach to enhance the teaching of movement amplitude skills. Our study provides insights for improving the design of immersive shared systems. It opens up new avenues for further research on the effectiveness of shared immersive virtual environments in supporting the teaching of technical skills.