@article{hou_confer:_2026,
	title = {{CONFER}: {Conflict}-{Aware} {Evidence} {Negotiation} for {Regime}-{Calibrated} {Weak} {Supervision} in {Multimodal} {Emotion} {Recognition}},
	abstract = {Multimodal emotion recognition often treats self-reported labels as reliable supervision while overlooking self-report unreliability and cross-modal conflict. We propose CONFER, a graph-based conflict-aware evidence negotiation framework for weakly supervised multimodal emotion recognition. CONFER represents each modality expert as a node with a predictive belief, boundary-based uncertainty, and runtime reliability estimated from historical out-of-fold performance and current-sample uncertainty. Uncertaintyaware compatibility and reliability-directed asymmetric edge weights govern iterative message-passing negotiation, followed by peer-supported prediction readout. Conflict reduction, residual disagreement, and mean modality uncertainty further characterize three regimes—Consensus, Dissent, and Ambiguity—for sample-specific weak-label calibration. We evaluate CONFER on AMIGOS, MAHNOB-HCI, and DEAP under subject-dependent 10-fold and strict leave-one-subjectout (LOSO) protocols. CONFER achieves competitive performance, reaching 0.873 accuracy on AMIGOS-V and 0.854 accuracy on MAHNOB-V under strict LOSO evaluation. Further analyses show larger negotiation gains on high-conflict samples and improved robustness to weak-label corruption, indicating that cross-modal conflict provides useful information for both directional modality coordination and supervisionreliability estimation.},
	language = {en},
	author = {Hou, Bojing and Li, Ruohao and Zhu, Yitong and Yu, Luwen and Wang, Yuyang},
	month = aug,
	year = {2026},
}

Multimodal emotion recognition often treats self-reported labels as reliable supervision while overlooking self-report unreliability and cross-modal conflict. We propose CONFER, a graph-based conflict-aware evidence negotiation framework for weakly supervised multimodal emotion recognition. CONFER represents each modality expert as a node with a predictive belief, boundary-based uncertainty, and runtime reliability estimated from historical out-of-fold performance and current-sample uncertainty. Uncertaintyaware compatibility and reliability-directed asymmetric edge weights govern iterative message-passing negotiation, followed by peer-supported prediction readout. Conflict reduction, residual disagreement, and mean modality uncertainty further characterize three regimes—Consensus, Dissent, and Ambiguity—for sample-specific weak-label calibration. We evaluate CONFER on AMIGOS, MAHNOB-HCI, and DEAP under subject-dependent 10-fold and strict leave-one-subjectout (LOSO) protocols. CONFER achieves competitive performance, reaching 0.873 accuracy on AMIGOS-V and 0.854 accuracy on MAHNOB-V under strict LOSO evaluation. Further analyses show larger negotiation gains on high-conflict samples and improved robustness to weak-label corruption, indicating that cross-modal conflict provides useful information for both directional modality coordination and supervisionreliability estimation.

@misc{hou_biokd:_2026,
	title = {{BioKD}: {Selective} {Physiology}-to-{Video} {Knowledge} {Distillation} via {Reliability} {Gate} for {Emotion} {Recognition}},
	copyright = {Creative Commons Attribution 4.0 International},
	shorttitle = {{BioKD}},
	url = {https://arxiv.org/abs/2608.06023},
	doi = {10.48550/ARXIV.2608.06023},
	abstract = {To address the limitations of video-based emotion recognition under ambiguous or socially masked behavioral cues, as well as the poor deployability of physiological signals, this paper proposes a reliability-aware physiology-to-video knowledge distillation framework, termed BioKD. The proposed framework leverages physiological signals as privileged information during training to guide a video-based student model in learning deep affective representations, while relying solely on non-intrusive video inputs at inference time. To cope with the high noise and instability of physiological teacher supervision caused by inter-subject variability, signal artifacts, and temporal inconsistency, BioKD incorporates a sample-wise reliability-aware gating mechanism together with a progressive distillation strategy. By adaptively regulating the strength of knowledge transfer, the framework suppresses negative transfer induced by unreliable physiological supervision and enables more stable cross-modal distillation. Experiments on DEAP and AMIGOS show that BioKD consistently outperforms representative baselines under both trial-wise and subject-wise evaluation protocols for valence and arousal recognition. For example, BioKD achieves 68.01{\textbackslash}\% on DEAP (trial-wise arousal) and 65.29{\textbackslash}\% under the more challenging subject-wise setting, demonstrating improved performance under a subject-independent evaluation setting. Further analyses show that BioKD effectively mitigates overconfident teacher errors and outperforms an entropy-only weighting strategy, confirming the importance of explicitly modeling supervision reliability. In addition, BioKD introduces no additional inference-time overhead relative to the same video student architecture and removes the need for physiological sensing and multimodal synchronization.},
	urldate = {2026-08-09},
	publisher = {arXiv},
	author = {Hou, Bojing and Li, Ruohao and Zhu, Yitong and Liu, Hongjun and Yu, Luwen and Wang, Yuyang},
	month = aug,
	year = {2026},
	note = {Version Number: 1},
}

To address the limitations of video-based emotion recognition under ambiguous or socially masked behavioral cues, as well as the poor deployability of physiological signals, this paper proposes a reliability-aware physiology-to-video knowledge distillation framework, termed BioKD. The proposed framework leverages physiological signals as privileged information during training to guide a video-based student model in learning deep affective representations, while relying solely on non-intrusive video inputs at inference time. To cope with the high noise and instability of physiological teacher supervision caused by inter-subject variability, signal artifacts, and temporal inconsistency, BioKD incorporates a sample-wise reliability-aware gating mechanism together with a progressive distillation strategy. By adaptively regulating the strength of knowledge transfer, the framework suppresses negative transfer induced by unreliable physiological supervision and enables more stable cross-modal distillation. Experiments on DEAP and AMIGOS show that BioKD consistently outperforms representative baselines under both trial-wise and subject-wise evaluation protocols for valence and arousal recognition. For example, BioKD achieves 68.01\% on DEAP (trial-wise arousal) and 65.29\% under the more challenging subject-wise setting, demonstrating improved performance under a subject-independent evaluation setting. Further analyses show that BioKD effectively mitigates overconfident teacher errors and outperforms an entropy-only weighting strategy, confirming the importance of explicitly modeling supervision reliability. In addition, BioKD introduces no additional inference-time overhead relative to the same video student architecture and removes the need for physiological sensing and multimodal synchronization.

@article{li_adaptation_2026,
	title = {From {Adaptation} to {Intelligence}: {A} {Systematic} {Review} of {Data}, {Strategies}, and {Impact} in {Personalized} {VR}},
	volume = {Accepted},
	doi = {10.1080/10447318.2026.2712140},
	journal = {International Journal of Human-Computer Interaction},
	author = {Li, Tangyao and Zhu, Yitong and Liang, Hai-Ning and Wang, Yuyang},
	month = aug,
	year = {2026},
}

@article{bai_aligning_2026,
	title = {Aligning {Gamification} with {Learner} {Motivation}: {Insights} from {VR}-{Based} {Learning} {Tasks}},
	volume = {32},
	copyright = {https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html},
	issn = {1077-2626, 1941-0506, 2160-9306},
	shorttitle = {Aligning {Gamification} with {Learner} {Motivation}},
	url = {https://ieeexplore.ieee.org/document/11479635/},
	doi = {10.1109/TVCG.2026.3680716},
	abstract = {Virtual reality is increasingly adopted in education for its potential to create immersive, interactive learning experiences. Among various instructional approaches, gamification has emerged as a promising strategy to enhance learner engagement and motivation. While VR naturally provides a platform for game-based learning, current applications often overlook the underlying motivational mechanisms that influence learner behavior in these e nvironments. Building on Self-Determination Theory, we examined how gamification and motivational framing influence VR le arning. We designed a VR system featuring both gamified and non-gamified versions of two learning tasks—one culturally expressive (batik-based task) and one technically focused (code-based task)—to reflect different motivational framings. Our findings show that gamification selectively influenced intrinsic motivation components and played a dominant role in shaping learner satisfaction. A trend-level interaction suggested that combining extrinsic framing with gamification may increase satisfaction. Gamification improved overall experience and engagement, even though it elevated perceived workload. These results highlight the importance of aligning gamification strategies with learner motivation and underscore the need for more dynamic approaches to capture motivational processes in VR, offering design insights for more effective and psychologically attuned VR learning environments.},
	language = {en},
	number = {5},
	urldate = {2026-06-09},
	journal = {IEEE Transactions on Visualization and Computer Graphics},
	author = {Bai, Yuxuan and Wang, Yu-Zhu and Hui, Pan and Wang, Yuyang},
	month = may,
	year = {2026},
	pages = {3411--3421},
}

Virtual reality is increasingly adopted in education for its potential to create immersive, interactive learning experiences. Among various instructional approaches, gamification has emerged as a promising strategy to enhance learner engagement and motivation. While VR naturally provides a platform for game-based learning, current applications often overlook the underlying motivational mechanisms that influence learner behavior in these e nvironments. Building on Self-Determination Theory, we examined how gamification and motivational framing influence VR le arning. We designed a VR system featuring both gamified and non-gamified versions of two learning tasks—one culturally expressive (batik-based task) and one technically focused (code-based task)—to reflect different motivational framings. Our findings show that gamification selectively influenced intrinsic motivation components and played a dominant role in shaping learner satisfaction. A trend-level interaction suggested that combining extrinsic framing with gamification may increase satisfaction. Gamification improved overall experience and engagement, even though it elevated perceived workload. These results highlight the importance of aligning gamification strategies with learner motivation and underscore the need for more dynamic approaches to capture motivational processes in VR, offering design insights for more effective and psychologically attuned VR learning environments.

@incollection{wang_virtual_2026,
	address = {Cham},
	title = {Virtual {Reality}},
	isbn = {978-3-032-03295-9 978-3-032-03296-6},
	url = {https://link.springer.com/10.1007/978-3-032-03296-6_2},
	doi = {10.1007/978-3-032-03296-6_2},
	language = {en},
	urldate = {2026-06-09},
	booktitle = {Handbook of the {Metaverse}},
	publisher = {Springer Nature Switzerland},
	author = {Wang, Yuyang and Chardonnet, Jean-Rémy},
	year = {2026},
	pages = {55--102},
}

@inproceedings{zhu_qa-moe:_2026,
	address = {San Diego, California, United States},
	title = {{QA}-{MoE}: {Towards} a {Continuous} {Reliability} {Spectrum} with {Quality}-{Aware} {Mixture} of {Experts} for {Robust} {Multimodal} {Sentiment} {Analysis}},
	isbn = {979-8-89176-390-6},
	url = {https://aclanthology.org/2026.acl-long.1461/},
	abstract = {Multimodal Sentiment Analysis (MSA) aims to infer human sentiment from textual, acoustic, and visual signals. In real-world scenarios, however, multimodal inputs are often compromised by dynamic noise or modality missingness. Existing methods typically treat these imperfections as discrete cases or assume fixed corruption ratios, which limits their adaptability to continuously varying reliability conditions. To address this, we first introduce a Continuous Reliability Spectrum to unify missingness and quality degradation into a single framework. Building on this, we propose QA-MoE, a Quality-Aware Mixture-of-Experts framework that quantifies modality reliability via self-supervised aleatoric uncertainty. This mechanism explicitly guides expert routing, enabling the model to suppress error propagation from unreliable signals while preserving task-relevant information. Extensive experiments indicate that QA-MoE achieves competitive or state-of-the-art performance across diverse degradation scenarios and exhibits a promising One-Checkpoint-for-All property in practice.},
	booktitle = {Proceedings of the 64th {Annual} {Meeting} of the {Association} for {Computational} {Linguistics} ({Volume} 1: {Long} {Papers})},
	publisher = {Association for Computational Linguistics},
	author = {Zhu, Yitong and Jiang, Yuxuan and Jiang, Guanxuan and Hou, Bojing and Zhou, Peng Yuan and Lin, Ge and Wang, Yuyang},
	editor = {Liakata, Maria and Moreira, Viviane P. and Zhang, Jiajun and Jurgens, David},
	month = jul,
	year = {2026},
	pages = {31689--31703},
}

Multimodal Sentiment Analysis (MSA) aims to infer human sentiment from textual, acoustic, and visual signals. In real-world scenarios, however, multimodal inputs are often compromised by dynamic noise or modality missingness. Existing methods typically treat these imperfections as discrete cases or assume fixed corruption ratios, which limits their adaptability to continuously varying reliability conditions. To address this, we first introduce a Continuous Reliability Spectrum to unify missingness and quality degradation into a single framework. Building on this, we propose QA-MoE, a Quality-Aware Mixture-of-Experts framework that quantifies modality reliability via self-supervised aleatoric uncertainty. This mechanism explicitly guides expert routing, enabling the model to suppress error propagation from unreliable signals while preserving task-relevant information. Extensive experiments indicate that QA-MoE achieves competitive or state-of-the-art performance across diverse degradation scenarios and exhibits a promising One-Checkpoint-for-All property in practice.

@inproceedings{fang_guideme:_2026,
	address = {Barcelona Spain},
	title = {{GuideMe}: {A} {VLM}-{Based} {System} {Assisting} {Independent} {Smartphone} {Learning} for {Older} {Adults}},
	isbn = {979-8-4007-2278-3},
	shorttitle = {{GuideMe}},
	url = {https://dl.acm.org/doi/10.1145/3772318.3791448},
	doi = {10.1145/3772318.3791448},
	language = {en},
	urldate = {2026-06-09},
	booktitle = {Proceedings of the 2026 {CHI} {Conference} on {Human} {Factors} in {Computing} {Systems}},
	publisher = {ACM},
	author = {Fang, Kairong and Zhang, Jiesi and Ni, Shi-Ting and Hui, Pan and Wang, Yuyang},
	month = apr,
	year = {2026},
	pages = {1--19},
}

@article{xu_multimodal_2026,
	title = {Multimodal colour modulation for cognitive enhancement in intelligent rehabilitation: a systematic review and translational guidance},
	issn = {1743-0003},
	shorttitle = {Multimodal colour modulation for cognitive enhancement in intelligent rehabilitation},
	url = {https://link.springer.com/10.1186/s12984-026-01991-y},
	doi = {10.1186/s12984-026-01991-y},
	abstract = {Abstract
            
              Background
              Colour modulation has been investigated as a non-invasive means of supporting cognitive performance, yet translation to rehabilitation-oriented applications has been constrained by fragmented evidence, inconsistent parameterisation, and heterogeneous assessment strategies.
            
            
              Main body
              This systematic review synthesises 75 studies published between 2011 and 2025 through a Colour-Modality-Cognition framework that links stimulus design, measurement strategy, and cognitive goals. Across the evidence base, a graded and context-dependent pattern emerged. Contrast-related variables, particularly lightness and colour difference, showed more consistent support for baseline legibility and task reachability, whereas hue- and chroma-related effects were more conditional on task family, visual context, implementation locus, and the baseline contrast conditions already in place. Across several studies, efficiency-related changes, including shorter search times or reduced cortical load, were more readily captured by ocular and neurophysiological measures than by behavioural outcomes alone, especially where behavioural performance approached ceiling levels. However, the literature rarely assessed post-training retention, transfer, or other longer-term rehabilitation-relevant outcomes.
            
            
              Conclusion
              Current evidence supports a more explicit framework for the design and evaluation of rehabilitation-oriented colour interventions, particularly in relation to contrast-secured visual design and the use of multimodal assessment. However, implications for adaptive rehabilitation systems remain preliminary because most available evidence derives from controlled, predominantly non-clinical task settings rather than direct rehabilitation deployment.},
	language = {en},
	urldate = {2026-06-09},
	journal = {Journal of NeuroEngineering and Rehabilitation},
	author = {Xu, Lina and Wang, Xingkai and Li, Wenan and Jiang, Yuyang and Cheng, Chen and Li, Zhenhong and Wang, Yuyang and Yu, Luwen},
	month = may,
	year = {2026},
}

Abstract Background Colour modulation has been investigated as a non-invasive means of supporting cognitive performance, yet translation to rehabilitation-oriented applications has been constrained by fragmented evidence, inconsistent parameterisation, and heterogeneous assessment strategies. Main body This systematic review synthesises 75 studies published between 2011 and 2025 through a Colour-Modality-Cognition framework that links stimulus design, measurement strategy, and cognitive goals. Across the evidence base, a graded and context-dependent pattern emerged. Contrast-related variables, particularly lightness and colour difference, showed more consistent support for baseline legibility and task reachability, whereas hue- and chroma-related effects were more conditional on task family, visual context, implementation locus, and the baseline contrast conditions already in place. Across several studies, efficiency-related changes, including shorter search times or reduced cortical load, were more readily captured by ocular and neurophysiological measures than by behavioural outcomes alone, especially where behavioural performance approached ceiling levels. However, the literature rarely assessed post-training retention, transfer, or other longer-term rehabilitation-relevant outcomes. Conclusion Current evidence supports a more explicit framework for the design and evaluation of rehabilitation-oriented colour interventions, particularly in relation to contrast-secured visual design and the use of multimodal assessment. However, implications for adaptive rehabilitation systems remain preliminary because most available evidence derives from controlled, predominantly non-clinical task settings rather than direct rehabilitation deployment.

@article{zhu_flow-aware_2026,
	title = {Flow-{Aware} {Diffusion} for {Real}-{Time} {VR} {Restoration}: {Mitigating} {Cybersickness} with {Enhanced} {Spatiotemporal} {Coherence}},
	copyright = {https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html},
	issn = {1077-2626, 1941-0506, 2160-9306},
	shorttitle = {Flow-{Aware} {Diffusion} for {Real}-{Time} {VR} {Restoration}},
	url = {https://ieeexplore.ieee.org/document/11535293/},
	doi = {10.1109/TVCG.2026.3697013},
	abstract = {Cybersickness remains a critical barrier to the widespread adoption of Virtual Reality (VR), particularly in scenarios involving intense or artificial motion cues. Among the key contributors is excessive optical flow—perceived visual motion that, when unmatched by vestibular input, leads to sensory conflict and discomfort. While previous efforts have explored geometric or hardware-based mitigation strategies, such methods often rely on predefined scene structures, manual tuning, or intrusive equipment. In this work, we propose U-MAD, a lightweight, real-time, AI-based solution that suppresses perceptually disruptive optical flow directly at the image level. Unlike prior handcrafted approaches, this method learns to attenuate high-intensity motion patterns from rendered frames without requiring mesh-level editing or scene-specific adaptation. Designed as a plug-and-play module, U-MAD integrates seamlessly into existing VR pipelines and generalizes well to procedurally generated environments. The experiments show that U-MAD consistently reduces average optical flow and enhances temporal stability across diverse scenes. A user study further supports the finding that reducing visual motion defects can improve perceptual comfort and alleviate cybersickness symptoms. These findings demonstrate that perceptually guided modulation of optical flow provides an effective and scalable approach to creating more user-friendly immersive experiences. The code will be released at https://github.com/XXXXX (upon publication).},
	language = {en},
	urldate = {2026-06-09},
	journal = {IEEE Transactions on Visualization and Computer Graphics},
	author = {Zhu, Yitong and Jiang, Guanxuan and Liang, Zhuowen and Wang, Yuyang},
	year = {2026},
	pages = {1--15},
}

Cybersickness remains a critical barrier to the widespread adoption of Virtual Reality (VR), particularly in scenarios involving intense or artificial motion cues. Among the key contributors is excessive optical flow—perceived visual motion that, when unmatched by vestibular input, leads to sensory conflict and discomfort. While previous efforts have explored geometric or hardware-based mitigation strategies, such methods often rely on predefined scene structures, manual tuning, or intrusive equipment. In this work, we propose U-MAD, a lightweight, real-time, AI-based solution that suppresses perceptually disruptive optical flow directly at the image level. Unlike prior handcrafted approaches, this method learns to attenuate high-intensity motion patterns from rendered frames without requiring mesh-level editing or scene-specific adaptation. Designed as a plug-and-play module, U-MAD integrates seamlessly into existing VR pipelines and generalizes well to procedurally generated environments. The experiments show that U-MAD consistently reduces average optical flow and enhances temporal stability across diverse scenes. A user study further supports the finding that reducing visual motion defects can improve perceptual comfort and alleviate cybersickness symptoms. These findings demonstrate that perceptually guided modulation of optical flow provides an effective and scalable approach to creating more user-friendly immersive experiences. The code will be released at https://github.com/XXXXX (upon publication).

@article{jiang_when_2026,
	title = {When trust collides: {Exploring} human-{LLM} cooperation intention through the prisoner’s dilemma},
	volume = {209},
	issn = {10715819},
	shorttitle = {When trust collides},
	url = {https://linkinghub.elsevier.com/retrieve/pii/S1071581926000157},
	doi = {10.1016/j.ijhcs.2026.103740},
	abstract = {Large language models (LLMs) are advancing rapidly, equipping artificial intelligence (AI) agents with stronger reasoning and decision-making capacities. AI agents are moving from passive tools to more adaptive collaborators. However, the black-box nature and hallucinations of LLMs create uncertainties about their outputs, which may impact user trust in scenarios involving human-LLM collaboration and, in turn, lead to inconsistent outcomes in mixed-motive scenarios. To investigate how humans adjust their cooperation intentions in response to LLM unpredictability, we conducted repeated Prisoner’s Dilemma games with 30 participants (15 males, 15 females) interacting with LLM agents with different declared identities. Results revealed that an agent’s declared identity, the user’s gender, and their interaction influenced cooperation intentions. The semi-structured interviews further showed that these effects were mediated by gender differences in perceived agent identity and trustworthiness. By moving beyond the predominantly cooperative settings, this study uncovers robust interactions between agent identity and user gender in mixed-motive scenarios. These findings offer practical insights for developing more trustworthy AI systems for real-world problems.},
	language = {en},
	urldate = {2026-02-23},
	journal = {International Journal of Human-Computer Studies},
	author = {Jiang, Guanxuan and Yang, Shirao and Wang, Yuyang and Hui, Pan},
	month = feb,
	year = {2026},
	pages = {103740},
}

Large language models (LLMs) are advancing rapidly, equipping artificial intelligence (AI) agents with stronger reasoning and decision-making capacities. AI agents are moving from passive tools to more adaptive collaborators. However, the black-box nature and hallucinations of LLMs create uncertainties about their outputs, which may impact user trust in scenarios involving human-LLM collaboration and, in turn, lead to inconsistent outcomes in mixed-motive scenarios. To investigate how humans adjust their cooperation intentions in response to LLM unpredictability, we conducted repeated Prisoner’s Dilemma games with 30 participants (15 males, 15 females) interacting with LLM agents with different declared identities. Results revealed that an agent’s declared identity, the user’s gender, and their interaction influenced cooperation intentions. The semi-structured interviews further showed that these effects were mediated by gender differences in perceived agent identity and trustworthiness. By moving beyond the predominantly cooperative settings, this study uncovers robust interactions between agent identity and user gender in mixed-motive scenarios. These findings offer practical insights for developing more trustworthy AI systems for real-world problems.

@inproceedings{yu_vrtalk:_2025,
	address = {Daejeon, Korea, Republic of},
	title = {{VRtalk}: {Real}-{Time} {Interactive} {Intelligent} {Anime} {Avatars} in {Virtual} {Reality}},
	copyright = {https://doi.org/10.15223/policy-029},
	isbn = {979-8-3315-8761-1},
	shorttitle = {{VRtalk}},
	url = {https://ieeexplore.ieee.org/document/11220364/},
	doi = {10.1109/ISMAR67309.2025.00125},
	abstract = {The convergence of virtual reality live streaming and AI-driven avatars has emerged as a significant technological trend. However, current integration attempts remain in the proof-of-concept stage, with the primary challenge of automatic interaction system establishment. To build interactive intelligence anime avatars within VR frameworks, we have developed a multimodal interaction architecture centered on dialogue agents, realizing comprehensive understanding, reasoning, and response. Our approach 1).proposes high granularity explicit-implicit understanding and a dual-center switchable reasoning mechanism to support flexible responses. 2).innovates a dual-source animation mechanism for co-speech face-body visualization and a textual command module for supervising crossmodal animation, and 3).enhances expressiveness through mapping persona, content, voice, and motion to anime style. Experimental results demonstrate the state-of-the-art performance of VRtalk, highlighting its practical significance and future potential.},
	language = {en},
	urldate = {2026-06-09},
	booktitle = {2025 {IEEE} {International} {Symposium} on {Mixed} and {Augmented} {Reality} ({ISMAR})},
	publisher = {IEEE},
	author = {Yu, Yuan and Xu, Chunlei and Yang, Shirao and Cao, Yu and Wang, Yuyang and Lee, Boon Giin},
	month = oct,
	year = {2025},
	pages = {1191--1201},
}

The convergence of virtual reality live streaming and AI-driven avatars has emerged as a significant technological trend. However, current integration attempts remain in the proof-of-concept stage, with the primary challenge of automatic interaction system establishment. To build interactive intelligence anime avatars within VR frameworks, we have developed a multimodal interaction architecture centered on dialogue agents, realizing comprehensive understanding, reasoning, and response. Our approach 1).proposes high granularity explicit-implicit understanding and a dual-center switchable reasoning mechanism to support flexible responses. 2).innovates a dual-source animation mechanism for co-speech face-body visualization and a textual command module for supervising crossmodal animation, and 3).enhances expressiveness through mapping persona, content, voice, and motion to anime style. Experimental results demonstrate the state-of-the-art performance of VRtalk, highlighting its practical significance and future potential.

@article{qin_identity_2025,
	title = {Identity, crimes, and law enforcement in the {Metaverse}},
	volume = {12},
	issn = {2662-9992},
	url = {https://www.nature.com/articles/s41599-024-04266-w},
	doi = {10.1057/s41599-024-04266-w},
	language = {en},
	number = {1},
	urldate = {2026-06-09},
	journal = {Humanities and Social Sciences Communications},
	author = {Qin, Hua Xuan and Wang, Yuyang and Hui, Pan},
	month = feb,
	year = {2025},
	pages = {194},
}

@article{li_balancing_2025,
	title = {Balancing {Exploration} and {Cybersickness}: {Investigating} {Curiosity}-{Driven} {Behavior} in {Virtual} {Environments}},
	volume = {55},
	copyright = {https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html},
	issn = {2168-2291, 2168-2305},
	shorttitle = {Balancing {Exploration} and {Cybersickness}},
	url = {https://ieeexplore.ieee.org/document/11151616/},
	doi = {10.1109/THMS.2025.3602125},
	abstract = {Virtual reality offers the opportunity for immersive exploration, yet it is often undermined by cybersickness. However, how individuals strike a balance between exploration and discomfort remains unclear. Existing method (e.g., reinforcement learning (RL)) often fail to fully capture the complexities of navigation and decision-making patterns. This study investigates how curiosity influences users’ navigation behavior, particularly how users strike a balance between exploration and discomfort. We propose curiosity as a key factor driving irrational decision-making and apply the free energy principle to model the relationship between curiosity and user behavior quantitatively. Our findings indicate that users generally adopt conservative strategies when navigating. Also, curiosity levels tend to rise when the virtual environment changes. These results illustrate the dynamic interplay between exploration and discomfort. In addition, it offers a new perspective on how curiosity drives behavior in immersive environments, providing a foundation for designing adaptive VR environments. Future research will further refine this model by incorporating additional psychological and environmental factors to improve prediction accuracy.},
	language = {en},
	number = {6},
	urldate = {2026-02-23},
	journal = {IEEE Transactions on Human-Machine Systems},
	author = {Li, Tangyao and Wang, Yuyang},
	month = dec,
	year = {2025},
	pages = {1043--1052},
}

Virtual reality offers the opportunity for immersive exploration, yet it is often undermined by cybersickness. However, how individuals strike a balance between exploration and discomfort remains unclear. Existing method (e.g., reinforcement learning (RL)) often fail to fully capture the complexities of navigation and decision-making patterns. This study investigates how curiosity influences users’ navigation behavior, particularly how users strike a balance between exploration and discomfort. We propose curiosity as a key factor driving irrational decision-making and apply the free energy principle to model the relationship between curiosity and user behavior quantitatively. Our findings indicate that users generally adopt conservative strategies when navigating. Also, curiosity levels tend to rise when the virtual environment changes. These results illustrate the dynamic interplay between exploration and discomfort. In addition, it offers a new perspective on how curiosity drives behavior in immersive environments, providing a foundation for designing adaptive VR environments. Future research will further refine this model by incorporating additional psychological and environmental factors to improve prediction accuracy.

@inproceedings{zhu_towards_2025,
	address = {Dublin Ireland},
	title = {Towards {Consumer}-{Grade} {Cybersickness} {Prediction}: {Multi}-{Model} {Alignment} for {Real}-{Time} {Vision}-{Only} {Inference}},
	isbn = {979-8-4007-2035-2},
	shorttitle = {Towards {Consumer}-{Grade} {Cybersickness} {Prediction}},
	url = {https://dl.acm.org/doi/10.1145/3746027.3755115},
	doi = {10.1145/3746027.3755115},
	abstract = {Cybersickness remains a major obstacle to the widespread adoption of immersive virtual reality (VR), particularly in consumer-grade environments. While prior methods rely on invasive signals such as electroencephalography (EEG) for high predictive accuracy, these approaches require specialized hardware and are impractical for real-world applications. In this work, we propose a scalable, deployable framework for personalized cybersickness prediction leveraging only non-invasive signals readily available from commercial VR headsets, including head motion, eye tracking, and physiological responses. Our model employs a modality-speci�c graph neural network enhanced with a Di�erence Attention Module to extract temporal-spatial embeddings capturing dynamic changes across modalities. A cross-modal alignment module jointly trains the video encoder to learn personalized traits by aligning video features with sensor-derived representations. Consequently, the model accurately predicts individual cybersickness using only video input during inference. Experimental results show our model achieves 88.4\% accuracy, closely matching EEG-based approaches (89.16\%), while reducing deployment complexity. With an average inference latency of 90ms, our framework supports real-time applications, ideal for integration into consumer-grade VR platforms without compromising personalization or performance. The code will be relesed at https://github.com/U235-Aurora/PTGNN.},
	language = {en},
	urldate = {2026-02-23},
	booktitle = {Proceedings of the 33rd {ACM} {International} {Conference} on {Multimedia}},
	publisher = {ACM},
	author = {Zhu, Yitong and Liang, Zhuowen and Wu, Yiming and Li, Tangyao and Wang, Yuyang},
	month = oct,
	year = {2025},
	pages = {6859--6867},
}

Cybersickness remains a major obstacle to the widespread adoption of immersive virtual reality (VR), particularly in consumer-grade environments. While prior methods rely on invasive signals such as electroencephalography (EEG) for high predictive accuracy, these approaches require specialized hardware and are impractical for real-world applications. In this work, we propose a scalable, deployable framework for personalized cybersickness prediction leveraging only non-invasive signals readily available from commercial VR headsets, including head motion, eye tracking, and physiological responses. Our model employs a modality-speci�c graph neural network enhanced with a Di�erence Attention Module to extract temporal-spatial embeddings capturing dynamic changes across modalities. A cross-modal alignment module jointly trains the video encoder to learn personalized traits by aligning video features with sensor-derived representations. Consequently, the model accurately predicts individual cybersickness using only video input during inference. Experimental results show our model achieves 88.4% accuracy, closely matching EEG-based approaches (89.16%), while reducing deployment complexity. With an average inference latency of 90ms, our framework supports real-time applications, ideal for integration into consumer-grade VR platforms without compromising personalization or performance. The code will be relesed at https://github.com/U235-Aurora/PTGNN.

@misc{yang_interoperability_2024,
	title = {Interoperability of the {Metaverse}: {A} {Digital} {Ecosystem} {Perspective} {Review}},
	shorttitle = {Interoperability of the {Metaverse}},
	url = {https://www.ssrn.com/abstract=4929167},
	doi = {10.2139/ssrn.4929167},
	abstract = {The Metaverse, a pivotal element of the digital revolution, holds transformative potential for industries and lifestyles. Yet, skepticism persists, with concerns that enthusiasm may outstrip technological progress. Interoperability is a key obstacle, as highlighted by a CoinMarketCap report (February 2023) noting over 240 isolated Metaverse initiatives. Despite agreement on its importance, systematic research on interoperability remains scarce. This study bridges the gap through a systematic literature review, using content analysis on Web of Science and Scopus databases, identifying 74 relevant publications. Interoperability lacks a standardized definition, varying by context, while the Metaverse is broadly seen as a digital ecosystem. Urs Gasser’s framework for digital ecosystem interoperability—spanning technological, data, human, and institutional dimensions—guides our analysis. By applying this framework across three identified layers, we provide a comprehensive overview of Metaverse interoperability research, establishing benchmarks to advance scholarly exploration in this complex field.},
	language = {en},
	urldate = {2026-06-09},
	publisher = {SSRN},
	author = {Yang, Liang and Ni, Shi-Ting and Wang, Yuyang and Yu, Ao and Lee, Jyn-An and Hui, Pan},
	year = {2024},
}

The Metaverse, a pivotal element of the digital revolution, holds transformative potential for industries and lifestyles. Yet, skepticism persists, with concerns that enthusiasm may outstrip technological progress. Interoperability is a key obstacle, as highlighted by a CoinMarketCap report (February 2023) noting over 240 isolated Metaverse initiatives. Despite agreement on its importance, systematic research on interoperability remains scarce. This study bridges the gap through a systematic literature review, using content analysis on Web of Science and Scopus databases, identifying 74 relevant publications. Interoperability lacks a standardized definition, varying by context, while the Metaverse is broadly seen as a digital ecosystem. Urs Gasser’s framework for digital ecosystem interoperability—spanning technological, data, human, and institutional dimensions—guides our analysis. By applying this framework across three identified layers, we provide a comprehensive overview of Metaverse interoperability research, establishing benchmarks to advance scholarly exploration in this complex field.

@inproceedings{zhang_jump_2024,
	address = {Orlando, FL, USA},
	title = {Jump {Cut} {Effects} in {Cinematic} {Virtual} {Reality}: {Editing} with the 30-degree {Rule} and 180-degree {Rule}},
	copyright = {https://doi.org/10.15223/policy-029},
	isbn = {979-8-3503-7402-5},
	shorttitle = {Jump {Cut} {Effects} in {Cinematic} {Virtual} {Reality}},
	url = {https://ieeexplore.ieee.org/document/10494084/},
	doi = {10.1109/VR58804.2024.00029},
	abstract = {Virtual reality (VR) is an immersive medium that offers users a unique opportunity to experience a digital environment realistically. As the demand for VR content continues to grow, the importance of effective VR editing techniques becomes increasingly apparent. This paper is a pioneering work investigating the effects of jump cuts on the viewer’s sense of presence, viewing experience, and edit quality in cinematic VR. Specifically, this work focuses on using the 30-degree and 180-degree rules in VR editing to minimize the adverse effects of jump cuts. We conducted a user study with thirteen participants, who watched nine different VR edits and completed a survey for each edited video. Our results indicate that employing the 30-degree and 180-degree rules in VR editing can significantly improve the sense of presence, viewing experience, and edit quality while mitigating the negative effects of jump cuts. We provide valuable insights for VR content creators and editors to achieve more effective and immersive VR experiences.},
	language = {en},
	urldate = {2026-06-09},
	booktitle = {2024 {IEEE} {Conference} {Virtual} {Reality} and {3D} {User} {Interfaces} ({VR})},
	publisher = {IEEE},
	author = {Zhang, Junjie and Lee, Lik-Hang and Wang, Yuyang and Jin, Shan and Fei, Dan-Lu and Hui, Pan},
	month = mar,
	year = {2024},
	pages = {51--60},
}

Virtual reality (VR) is an immersive medium that offers users a unique opportunity to experience a digital environment realistically. As the demand for VR content continues to grow, the importance of effective VR editing techniques becomes increasingly apparent. This paper is a pioneering work investigating the effects of jump cuts on the viewer’s sense of presence, viewing experience, and edit quality in cinematic VR. Specifically, this work focuses on using the 30-degree and 180-degree rules in VR editing to minimize the adverse effects of jump cuts. We conducted a user study with thirteen participants, who watched nine different VR edits and completed a survey for each edited video. Our results indicate that employing the 30-degree and 180-degree rules in VR editing can significantly improve the sense of presence, viewing experience, and edit quality while mitigating the negative effects of jump cuts. We provide valuable insights for VR content creators and editors to achieve more effective and immersive VR experiences.

@inproceedings{liu_omnicolor:_2024,
	address = {Yokohama, Japan},
	title = {{OmniColor}: {A} {Global} {Camera} {Pose} {Optimization} {Approach} of {LiDAR}-{360Camera} {Fusion} for {Colorizing} {Point} {Clouds}},
	copyright = {https://doi.org/10.15223/policy-029},
	isbn = {979-8-3503-8457-4},
	shorttitle = {{OmniColor}},
	url = {https://ieeexplore.ieee.org/document/10610292/},
	doi = {10.1109/ICRA57147.2024.10610292},
	abstract = {A Colored point cloud, as a simple and efficient 3D representation, has many advantages in various fields, including robotic navigation and scene reconstruction. This representation is now commonly used in 3D reconstruction tasks relying on cameras and LiDARs. However, fusing data from these two types of sensors is poorly performed in many existing frameworks, leading to unsatisfactory mapping results, mainly due to inaccurate camera poses. This paper presents OmniColor, a novel and efficient algorithm to colorize point clouds using an independent 360-degree camera. Given a LiDARbased point cloud and a sequence of panorama images with initial coarse camera poses, our objective is to jointly optimize the poses of all frames for mapping images onto geometric reconstructions. Our pipeline works in an off-the-shelf manner that does not require any feature extraction or matching process. Instead, we find optimal poses by directly maximizing the photometric consistency of LiDAR maps. In experiments, we show that our method can overcome the severe visual distortion of omnidirectional images and greatly benefit from the wide field of view (FOV) of 360-degree cameras to reconstruct various scenarios with accuracy and stability. The code will be released at https://github.com/liubonan123/OmniColor/.},
	language = {en},
	urldate = {2026-06-09},
	booktitle = {2024 {IEEE} {International} {Conference} on {Robotics} and {Automation} ({ICRA})},
	publisher = {IEEE},
	author = {Liu, Bonan and Zhao, Guoyang and Jiao, Jianhao and Cai, Guang and Li, Chengyang and Yin, Handi and Wang, Yuyang and Liu, Ming and Hui, Pan},
	month = may,
	year = {2024},
	pages = {6396--6402},
}

A Colored point cloud, as a simple and efficient 3D representation, has many advantages in various fields, including robotic navigation and scene reconstruction. This representation is now commonly used in 3D reconstruction tasks relying on cameras and LiDARs. However, fusing data from these two types of sensors is poorly performed in many existing frameworks, leading to unsatisfactory mapping results, mainly due to inaccurate camera poses. This paper presents OmniColor, a novel and efficient algorithm to colorize point clouds using an independent 360-degree camera. Given a LiDARbased point cloud and a sequence of panorama images with initial coarse camera poses, our objective is to jointly optimize the poses of all frames for mapping images onto geometric reconstructions. Our pipeline works in an off-the-shelf manner that does not require any feature extraction or matching process. Instead, we find optimal poses by directly maximizing the photometric consistency of LiDAR maps. In experiments, we show that our method can overcome the severe visual distortion of omnidirectional images and greatly benefit from the wide field of view (FOV) of 360-degree cameras to reconstruct various scenarios with accuracy and stability. The code will be released at https://github.com/liubonan123/OmniColor/.

@article{zhu_study_2024,
	title = {A {Study} of {Partisan} {News} {Sharing} in the {Russian} {Invasion} of {Ukraine}},
	volume = {18},
	issn = {2334-0770, 2162-3449},
	url = {https://ojs.aaai.org/index.php/ICWSM/article/view/31430},
	doi = {10.1609/icwsm.v18i1.31430},
	abstract = {Since the Russian invasion of Ukraine, a large volume of biased and partisan news has been spread via social media platforms. As this may lead to wider societal issues, we argue that understanding how partisan news sharing impacts users’ communication is crucial for better governance of online communities. In this paper, we perform a measurement study of partisan news sharing. We aim to characterize the role of such sharing in influencing users’ communications. Our analysis covers an eight-month dataset across six Reddit communities related to the Russian invasion. We first perform an analysis of the temporal evolution of partisan news sharing. We confirm that the invasion stimulates discussion in the observed communities, accompanied by an increased volume of partisan news sharing. Next, we characterize users’ response to such sharing. We observe that partisan bias plays a role in narrowing its propagation. More biased media is less likely to be spread across multiple subreddits. However, we find that partisan news sharing attracts more users to engage in the discussion, by generating more comments. We then built a predictive model to identify users likely to spread partisan news. The prediction is challenging though, with 61.57\% accuracy on average. Our centrality analysis on the commenting network further indicates that the users who disseminate partisan news possess lower network influence in comparison to those who propagate neutral news.},
	language = {en},
	urldate = {2026-06-09},
	journal = {Proceedings of the International AAAI Conference on Web and Social Media},
	author = {Zhu, Yiming and Haq, Ehsan-Ul and Tyson, Gareth and Lee, Lik-Hang and Wang, Yuyang and Hui, Pan},
	month = may,
	year = {2024},
	pages = {1847--1858},
}

Since the Russian invasion of Ukraine, a large volume of biased and partisan news has been spread via social media platforms. As this may lead to wider societal issues, we argue that understanding how partisan news sharing impacts users’ communication is crucial for better governance of online communities. In this paper, we perform a measurement study of partisan news sharing. We aim to characterize the role of such sharing in influencing users’ communications. Our analysis covers an eight-month dataset across six Reddit communities related to the Russian invasion. We first perform an analysis of the temporal evolution of partisan news sharing. We confirm that the invasion stimulates discussion in the observed communities, accompanied by an increased volume of partisan news sharing. Next, we characterize users’ response to such sharing. We observe that partisan bias plays a role in narrowing its propagation. More biased media is less likely to be spread across multiple subreddits. However, we find that partisan news sharing attracts more users to engage in the discussion, by generating more comments. We then built a predictive model to identify users likely to spread partisan news. The prediction is challenging though, with 61.57% accuracy on average. Our centrality analysis on the commenting network further indicates that the users who disseminate partisan news possess lower network influence in comparison to those who propagate neutral news.

@article{luo_using_2024,
	title = {Using a {Virtual} {Reality} {Interview} {Simulator} to {Explore} {Factors} {Influencing} {People}’s {Behavior}},
	volume = {28},
	issn = {1359-4338, 1434-9957},
	url = {https://link.springer.com/10.1007/s10055-023-00934-5},
	doi = {10.1007/s10055-023-00934-5},
	abstract = {Abstract
            
              Virtual reality interview simulator (VRIS) is an effective and valid tool that uses virtual reality technology to train people’s interview skills. Typically, it offers candidates prone to being very nervous during interviews the opportunity to practice interviews in a safe and manageable virtual environment and realistic settings, providing real-time feedback from a virtual interviewer on their performance. It helps interviewees improve their skills, reduce their fears, gain confidence, and minimize the cost and time associated with traditional interview preparation. Yet, the major anxiety-inducing elements remain unknown. During an interview, the anxiety levels, overall experience, and performance of interviewees might be affected by various circumstances. By analyzing electrodermal activity and questionnaire, we investigated the influence of five variables: (I)
              Realism
              ; (II)
              Question type
              ; (III)
              Interviewer attitude
              ; (IV)
              Timing
              ; and (V)
              Preparation
              . As such, an orthogonal design
              
                
                  \$\$L\_8(4{\textasciicircum}1 {\textbackslash}times 2{\textasciicircum}4)\$\$
                  
                    
                      
                        L
                        8
                      
                      
                        (
                        
                          4
                          1
                        
                        ×
                        
                          2
                          4
                        
                        )
                      
                    
                  
                
              
              with eight experiments (
              
                
                  \$\$O A\_8\$\$
                  
                    
                      O
                      
                        A
                        8
                      
                    
                  
                
              
              matrix) was implemented, in which 19 college students took part in the experiments. Considering the anxiety, overall experience, and performance of the interviewees, we found that
              Question type
              plays a major role; secondly,
              Realism
              ,
              Preparation
              , and
              Interviewer attitude
              all have middle influence; lastly,
              Timing
              has little to no impact. Specifically, professional interview questions elicited a greater degree of anxiety than personal ones among the categories of interview questions. This work contributes to our understanding of anxiety-stimulating factors during job interviews in virtual reality and provides cues for designing future VRIS.},
	language = {en},
	number = {1},
	urldate = {2024-03-15},
	journal = {Virtual Reality},
	author = {Luo, Xinyi and Wang, Yuyang and Lee, Lik-Hang and Xing, Zihan and Jin, Shan and Dong, Boya and Hu, Yuanyi and Chen, Zeming and Yan, Jing and Hui, Pan},
	month = mar,
	year = {2024},
	pages = {56},
}

Abstract Virtual reality interview simulator (VRIS) is an effective and valid tool that uses virtual reality technology to train people’s interview skills. Typically, it offers candidates prone to being very nervous during interviews the opportunity to practice interviews in a safe and manageable virtual environment and realistic settings, providing real-time feedback from a virtual interviewer on their performance. It helps interviewees improve their skills, reduce their fears, gain confidence, and minimize the cost and time associated with traditional interview preparation. Yet, the major anxiety-inducing elements remain unknown. During an interview, the anxiety levels, overall experience, and performance of interviewees might be affected by various circumstances. By analyzing electrodermal activity and questionnaire, we investigated the influence of five variables: (I) Realism ; (II) Question type ; (III) Interviewer attitude ; (IV) Timing ; and (V) Preparation . As such, an orthogonal design $$L_8(4\textasciicircum1 \times 2\textasciicircum4)$$ L 8 ( 4 1 × 2 4 ) with eight experiments ( $$O A_8$$ O A 8 matrix) was implemented, in which 19 college students took part in the experiments. Considering the anxiety, overall experience, and performance of the interviewees, we found that Question type plays a major role; secondly, Realism , Preparation , and Interviewer attitude all have middle influence; lastly, Timing has little to no impact. Specifically, professional interview questions elicited a greater degree of anxiety than personal ones among the categories of interview questions. This work contributes to our understanding of anxiety-stimulating factors during job interviews in virtual reality and provides cues for designing future VRIS.

@incollection{luo_modeling_2024,
	address = {Singapore},
	title = {Modeling {Online} {Adaptive} {Navigation} in {Virtual} {Environments} {Based} on {PID} {Control}},
	volume = {1964},
	isbn = {978-981-99-8140-3 978-981-99-8141-0},
	url = {https://link.springer.com/10.1007/978-981-99-8141-0_25},
	doi = {10.1007/978-981-99-8141-0_25},
	language = {en},
	urldate = {2024-03-10},
	booktitle = {Neural {Information} {Processing}},
	publisher = {Springer Nature Singapore},
	author = {Wang, Yuyang and Chardonnet, Jean-Rémy and Merienne, Frédéric},
	editor = {Luo, Biao and Cheng, Long and Wu, Zheng-Guang and Li, Hongyi and Li, Chaojie},
	year = {2024},
	note = {Series Title: Communications in Computer and Information Science},
	pages = {325--346},
}

@inproceedings{yin_text2vrscene:_2024,
	address = {Orlando, USA},
	title = {{Text2VRScene}: {Exploring} the {Paradigm} of {Automated} {Generation} {System} for {VR} {Experience} {From} the {Ground} {Up}},
	language = {en},
	booktitle = {2024 {IEEE} {Virtual} {Reality} and {3D} {User} {Interfaces} ({VR})},
	publisher = {IEEE},
	author = {Yin, Zhizhuo and Wang, Yuyang and Papatheodorou, Theodoros and Hui, Pan},
	month = mar,
	year = {2024},
}

@inproceedings{jin_development_2023,
	address = {Guangzhou China},
	title = {Development of an immersive simulator for improving student chemistry learning efficiency},
	isbn = {979-8-4007-0751-3},
	url = {https://dl.acm.org/doi/10.1145/3615522.3615535},
	doi = {10.1145/3615522.3615535},
	abstract = {Virtual reality (VR) technology has been used for educational purposes in different learning contents during teaching and training. VR could improve users’ learning efficiency and motivation to study abstract concepts. This work designed a VR environment for chemistry education to support computer-mediated hands-on exercises, including Self-propagating high-temperature synthesis (SHS) and Electrode sheet fabrication (ESF). In our evaluation with 39 participants who wore heart beat measurement wearables, we compared the students’ performances in hands-on chemistry tasks, either with or without score-keeping and time-sensitive conditions. Accordingly, we designed questionnaires reflecting sixteen qualitative aspects (e.g., content, perspicuity, and interaction) and perceived user workloads. The experimental results indicate participants’ preferences and attitudes in terms of efficiency and sense of safety. 94.87\% of participants reported that the learning simulator could improve learning efficiency, and 92.31\% of the participants indicated that it can improve their sense of safety. The results of the data analysis show that the different learning scenarios we simulated have positive significance. Our findings shed light on the quality and learning performance of operational skills for chemistry education.},
	language = {en},
	urldate = {2026-06-19},
	booktitle = {Proceedings of the 16th {International} {Symposium} on {Visual} {Information} {Communication} and {Interaction}},
	publisher = {ACM},
	author = {Jin, Shan and Wang, Yuyang and Lee, Lik-Hang and Luo, Xinyi and Hui, Pan},
	month = sep,
	year = {2023},
	pages = {1--8},
}

Virtual reality (VR) technology has been used for educational purposes in different learning contents during teaching and training. VR could improve users’ learning efficiency and motivation to study abstract concepts. This work designed a VR environment for chemistry education to support computer-mediated hands-on exercises, including Self-propagating high-temperature synthesis (SHS) and Electrode sheet fabrication (ESF). In our evaluation with 39 participants who wore heart beat measurement wearables, we compared the students’ performances in hands-on chemistry tasks, either with or without score-keeping and time-sensitive conditions. Accordingly, we designed questionnaires reflecting sixteen qualitative aspects (e.g., content, perspicuity, and interaction) and perceived user workloads. The experimental results indicate participants’ preferences and attitudes in terms of efficiency and sense of safety. 94.87% of participants reported that the learning simulator could improve learning efficiency, and 92.31% of the participants indicated that it can improve their sense of safety. The results of the data analysis show that the different learning scenarios we simulated have positive significance. Our findings shed light on the quality and learning performance of operational skills for chemistry education.

@inproceedings{li_deep_2023,
	address = {Sydney, Australia},
	title = {A {Deep} {Cybersickness} {Predictor} through {Kinematic} {Data} with {Encoded} {Physiological} {Representation}},
	copyright = {https://doi.org/10.15223/policy-029},
	isbn = {979-8-3503-2838-7},
	url = {https://ieeexplore.ieee.org/document/10316407/},
	doi = {10.1109/ISMAR59233.2023.00130},
	abstract = {Users would experience individually different sickness symptoms during or after navigating through an immersive virtual environment, generally known as cybersickness. Previous studies have predicted the severity of cybersickness based on physiological and/or kinematic data. However, compared with kinematic data, physiological data rely heavily on biosensors during the collection, which is inconvenient and limited to a few affordable VR devices. In this work, we proposed a deep neural network to predict cybersickness through kinematic data. We introduced the encoded physiological representation to characterize the individual susceptibility; therefore, the predictor could predict cybersickness only based on a user’s kinematic data without counting on biosensors. Fifty-three participants were recruited to attend the user study to collect multimodal data, including kinematic data (navigation speed, head tracking), physiological signals (e.g., electrodermal activity, heart rate), and Simulator Sickness Questionnaire (SSQ). The predictor achieved an accuracy of 97.8\% for cybersickness prediction by involving the pre-computed physiological representation to characterize individual differences, providing much convenience for the current cybersickness measurement.},
	language = {en},
	urldate = {2026-06-19},
	booktitle = {2023 {IEEE} {International} {Symposium} on {Mixed} and {Augmented} {Reality} ({ISMAR})},
	publisher = {IEEE},
	author = {Li, Ruichen and Wang, Yuyang and Yin, Handi and Chardonnet, Jean-Rémy and Hui, Pan},
	month = oct,
	year = {2023},
	pages = {1132--1141},
}

Users would experience individually different sickness symptoms during or after navigating through an immersive virtual environment, generally known as cybersickness. Previous studies have predicted the severity of cybersickness based on physiological and/or kinematic data. However, compared with kinematic data, physiological data rely heavily on biosensors during the collection, which is inconvenient and limited to a few affordable VR devices. In this work, we proposed a deep neural network to predict cybersickness through kinematic data. We introduced the encoded physiological representation to characterize the individual susceptibility; therefore, the predictor could predict cybersickness only based on a user’s kinematic data without counting on biosensors. Fifty-three participants were recruited to attend the user study to collect multimodal data, including kinematic data (navigation speed, head tracking), physiological signals (e.g., electrodermal activity, heart rate), and Simulator Sickness Questionnaire (SSQ). The predictor achieved an accuracy of 97.8% for cybersickness prediction by involving the pre-computed physiological representation to characterize individual differences, providing much convenience for the current cybersickness measurement.

@misc{wang_dataset_2023,
	title = {Dataset for predicting cybersickness from a virtual navigation task},
	url = {http://arxiv.org/abs/2303.13527},
	abstract = {This work presents a dataset collected to predict cybersickness in virtual reality environments. The data was collected from navigation tasks in a virtual environment designed to induce cybersickness. The dataset consists of many data points collected from diverse participants, including physiological responses (EDA and Heart Rate) and self-reported cybersickness symptoms. The paper will provide a detailed description of the dataset, including the arranged navigation task, the data collection procedures, and the data format. The dataset will serve as a valuable resource for researchers to develop and evaluate predictive models for cybersickness and will facilitate more research in cybersickness mitigation.},
	urldate = {2023-04-13},
	publisher = {arXiv},
	author = {Wang, Yuyang and Li, Ruichen and Chardonnet, Jean-Rémy and Hui, Pan},
	month = feb,
	year = {2023},
	note = {arXiv:2303.13527 [cs]},
}

This work presents a dataset collected to predict cybersickness in virtual reality environments. The data was collected from navigation tasks in a virtual environment designed to induce cybersickness. The dataset consists of many data points collected from diverse participants, including physiological responses (EDA and Heart Rate) and self-reported cybersickness symptoms. The paper will provide a detailed description of the dataset, including the arranged navigation task, the data collection procedures, and the data format. The dataset will serve as a valuable resource for researchers to develop and evaluate predictive models for cybersickness and will facilitate more research in cybersickness mitigation.

@inproceedings{wang_re-shaping_2022,
	address = {Bologna, Italy},
	title = {Re-shaping {Post}-{COVID}-19 {Teaching} and {Learning}: {A} {Blueprint} of {Virtual}-{Physical} {Blended} {Classrooms} in the {Metaverse} {Era}},
	copyright = {https://doi.org/10.15223/policy-029},
	isbn = {978-1-6654-8879-2},
	shorttitle = {Re-shaping {Post}-{COVID}-19 {Teaching} and {Learning}},
	url = {https://ieeexplore.ieee.org/document/9951355/},
	doi = {10.1109/ICDCSW56584.2022.00053},
	abstract = {During the COVID-19 pandemic, most countries have experienced some form of remote education through video conferencing software platforms. However, these software platforms fail to reduce immersion and replicate the classroom experience. The currently emerging Metaverse addresses many of such limitations by offering blended physical-digital environments. This paper aims to assess how the Metaverse can support and improve e-learning. We first survey the latest applications of blended environments in education and highlight the primary challenges and opportunities. Accordingly, we derive our proposal for a virtual-physical blended classroom configuration that brings students and teachers into a shared educational Metaverse. We focus on the system architecture of the Metaverse classroom to achieve real-time synchronization of a large number of participants and activities across physical (mixed reality classrooms) and virtual (remote VR platform) learning spaces. Our proposal attempts to transform the traditional physical classroom into virtual-physical cyberspace as a new social network of learners and educators connected at an unprecedented scale.},
	language = {en},
	urldate = {2026-06-09},
	booktitle = {2022 {IEEE} 42nd {International} {Conference} on {Distributed} {Computing} {Systems} {Workshops} ({ICDCSW})},
	publisher = {IEEE},
	author = {Wang, Yuyang and Lee, Lik-Hang and Braud, Tristan and Hui, Pan},
	month = jul,
	year = {2022},
	pages = {241--247},
}

During the COVID-19 pandemic, most countries have experienced some form of remote education through video conferencing software platforms. However, these software platforms fail to reduce immersion and replicate the classroom experience. The currently emerging Metaverse addresses many of such limitations by offering blended physical-digital environments. This paper aims to assess how the Metaverse can support and improve e-learning. We first survey the latest applications of blended environments in education and highlight the primary challenges and opportunities. Accordingly, we derive our proposal for a virtual-physical blended classroom configuration that brings students and teachers into a shared educational Metaverse. We focus on the system architecture of the Metaverse classroom to achieve real-time synchronization of a large number of participants and activities across physical (mixed reality classrooms) and virtual (remote VR platform) learning spaces. Our proposal attempts to transform the traditional physical classroom into virtual-physical cyberspace as a new social network of learners and educators connected at an unprecedented scale.

@article{hadadi_prediction_2022,
	title = {Prediction of cybersickness in virtual environments using topological data analysis and machine learning},
	volume = {3},
	issn = {2673-4192},
	url = {https://www.frontiersin.org/articles/10.3389/frvir.2022.973236/full},
	doi = {10.3389/frvir.2022.973236},
	abstract = {Recent significant progress in Virtual Reality (VR) applications and environments raised several challenges. They proved to have side effects on specific users, thus reducing the usability of the VR technology in some critical domains, such as flight and car simulators. One of the common side effects is cybersickness. Some significant commonly reported symptoms are nausea, oculomotor discomfort, and disorientation. To mitigate these symptoms and consequently improve the usability of VR systems, it is necessary to predict the incidence of cybersickness. This paper proposes a machine learning approach to VR’s cybersickness prediction based on physiological and subjective data. We investigated combinations of topological data analysis with a range of classifier algorithms and assessed classification performance. The highest performance of Topological Data Analysis (TDA) based methods was achieved in combination with SVMs with Gaussian RBF kernel, indicating that Gaussian RBF kernels provide embeddings of physiological time series data into spaces that are rich enough to capture the essential geometric features of this type of data. Comparing several combinations with feature descriptors for physiological time series, the performance of the TDA + SVM combination is in the top group, statistically being on par or outperforming more complex and less interpretable methods. Our results show that heart rate does not seem to correlate with cybersickness.},
	language = {en},
	urldate = {2026-06-09},
	journal = {Frontiers in Virtual Reality},
	author = {Hadadi, Azadeh and Guillet, Christophe and Chardonnet, Jean-Rémy and Langovoy, Mikhail and Wang, Yuyang and Ovtcharova, Jivka},
	month = oct,
	year = {2022},
	pages = {973236},
}

Recent significant progress in Virtual Reality (VR) applications and environments raised several challenges. They proved to have side effects on specific users, thus reducing the usability of the VR technology in some critical domains, such as flight and car simulators. One of the common side effects is cybersickness. Some significant commonly reported symptoms are nausea, oculomotor discomfort, and disorientation. To mitigate these symptoms and consequently improve the usability of VR systems, it is necessary to predict the incidence of cybersickness. This paper proposes a machine learning approach to VR’s cybersickness prediction based on physiological and subjective data. We investigated combinations of topological data analysis with a range of classifier algorithms and assessed classification performance. The highest performance of Topological Data Analysis (TDA) based methods was achieved in combination with SVMs with Gaussian RBF kernel, indicating that Gaussian RBF kernels provide embeddings of physiological time series data into spaces that are rich enough to capture the essential geometric features of this type of data. Comparing several combinations with feature descriptors for physiological time series, the performance of the TDA + SVM combination is in the top group, statistically being on par or outperforming more complex and less interpretable methods. Our results show that heart rate does not seem to correlate with cybersickness.

@article{wang_development_2021,
	title = {Development of a speed protector to optimize user experience in {3D} virtual environments},
	volume = {147},
	issn = {10715819},
	url = {https://linkinghub.elsevier.com/retrieve/pii/S1071581920301804},
	doi = {10.1016/j.ijhcs.2020.102578},
	abstract = {Virtual walking in virtual environments (VEs) requires locomotion interfaces, especially when the available physical environment is smaller than the virtual space due to virtual reality facilities limitations; many navigation approaches have been proposed according to different input conditions, target selection and speed selection. With current technologies, the virtual locomotion speed for most VR systems relies primarily on ratecontrol devices (e.g., joystick). The user has to manage manual adaptation of the speed, based on the size of the VE and personal preferences. However, this method cannot provide optimal speeds for locomotion as the user tends to change the speed involuntarily due to non-desired issues including collisions or simulator sickness; in this case, the user may have to adjust the speed frequently and unsmoothly, worsening the situation. Therefore, we designed a motion protector that can be embedded into the locomotion system to provide optimal speed profiles. The optimization process aims at minimizing the total jerk when the user translates from an initial position to a target, which is a common rule of the human motion model. In addition to minimization, we put constraints on speed, acceleration and jerk so that they do not exceed specific thresholds. The speed protector is formulated mathematically and solved analytically in order to provide a smooth navigation experience with a minimum jerk of trajectory. The assessment of the speed protector was conducted in a user study measuring user experience with a simulator sickness questionnaire, event-related skin conductance responses (ER-SCR), and a NASA-TLX questionnaire, showing that the designed speed protector can provide more natural and comfortable user experience with appropriate acceleration and jerk as it avoids abrupt speed profiles.},
	language = {en},
	urldate = {2026-06-09},
	journal = {International Journal of Human-Computer Studies},
	author = {Wang, Yuyang and Chardonnet, Jean-Rémy and Merienne, Frédéric},
	month = mar,
	year = {2021},
	pages = {102578},
}

Virtual walking in virtual environments (VEs) requires locomotion interfaces, especially when the available physical environment is smaller than the virtual space due to virtual reality facilities limitations; many navigation approaches have been proposed according to different input conditions, target selection and speed selection. With current technologies, the virtual locomotion speed for most VR systems relies primarily on ratecontrol devices (e.g., joystick). The user has to manage manual adaptation of the speed, based on the size of the VE and personal preferences. However, this method cannot provide optimal speeds for locomotion as the user tends to change the speed involuntarily due to non-desired issues including collisions or simulator sickness; in this case, the user may have to adjust the speed frequently and unsmoothly, worsening the situation. Therefore, we designed a motion protector that can be embedded into the locomotion system to provide optimal speed profiles. The optimization process aims at minimizing the total jerk when the user translates from an initial position to a target, which is a common rule of the human motion model. In addition to minimization, we put constraints on speed, acceleration and jerk so that they do not exceed specific thresholds. The speed protector is formulated mathematically and solved analytically in order to provide a smooth navigation experience with a minimum jerk of trajectory. The assessment of the speed protector was conducted in a user study measuring user experience with a simulator sickness questionnaire, event-related skin conductance responses (ER-SCR), and a NASA-TLX questionnaire, showing that the designed speed protector can provide more natural and comfortable user experience with appropriate acceleration and jerk as it avoids abrupt speed profiles.

@inproceedings{wang_using_2021,
	address = {Lisboa, Portugal},
	title = {Using {Fuzzy} {Logic} to {Involve} {Individual} {Differences} for {Predicting} {Cybersickness} during {VR} {Navigation}},
	copyright = {https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html},
	isbn = {978-1-6654-1838-6},
	url = {https://ieeexplore.ieee.org/document/9417785/},
	doi = {10.1109/VR50410.2021.00060},
	abstract = {Many studies have explored how individual differences can affect users’ susceptibility to cybersickness in a VR application. However, the lack of strategy to integrate the influence of each factor on cybersickness makes it difficult to utilize the results of existing research. Based on the fuzzy logic theory that can represent the effect of different factors as a single value containing integrated information, we developed two approaches including the knowledgebased Mamdani-type fuzzy inference system and the data-driven Adaptive neuro-fuzzy inference system (ANFIS) to involve three individual differences (Age, Gaming experience and Ethnicity). We correlated the corresponding outputs with the simulator sickness questionnaire (SSQ) scores in a simple navigation scenario. The correlation coefficients obtained through a 4-fold cross validation were found statistically significant with both fuzzy logic approaches, indicating their effectiveness to influence the occurrence and the level of cybersickness. Our work provides insights to establish customized experiences for VR navigation by involving individual differences.},
	language = {en},
	urldate = {2026-06-09},
	booktitle = {2021 {IEEE} {Virtual} {Reality} and {3D} {User} {Interfaces} ({VR})},
	publisher = {IEEE},
	author = {Wang, Yuyang and Chardonnet, Jean-Remy and Merienne, Frederic and Ovtcharova, Jivka},
	month = mar,
	year = {2021},
	pages = {373--381},
}

Many studies have explored how individual differences can affect users’ susceptibility to cybersickness in a VR application. However, the lack of strategy to integrate the influence of each factor on cybersickness makes it difficult to utilize the results of existing research. Based on the fuzzy logic theory that can represent the effect of different factors as a single value containing integrated information, we developed two approaches including the knowledgebased Mamdani-type fuzzy inference system and the data-driven Adaptive neuro-fuzzy inference system (ANFIS) to involve three individual differences (Age, Gaming experience and Ethnicity). We correlated the corresponding outputs with the simulator sickness questionnaire (SSQ) scores in a simple navigation scenario. The correlation coefficients obtained through a 4-fold cross validation were found statistically significant with both fuzzy logic approaches, indicating their effectiveness to influence the occurrence and the level of cybersickness. Our work provides insights to establish customized experiences for VR navigation by involving individual differences.

@article{wang_enhanced_2021,
	title = {Enhanced {Cognitive} {Workload} {Evaluation} in {3D} {Immersive} {Environments} with {TOPSIS} {Model}},
	volume = {147},
	issn = {10715819},
	url = {https://linkinghub.elsevier.com/retrieve/pii/S1071581920301749},
	doi = {10.1016/j.ijhcs.2020.102572},
	language = {en},
	urldate = {2024-03-10},
	journal = {International Journal of Human-Computer Studies},
	author = {Wang, Yuyang and Chardonnet, Jean-Rémy and Merienne, Frédéric},
	month = mar,
	year = {2021},
	pages = {102572},
}

@inproceedings{wang_vr_2019,
	address = {Osaka, Japan},
	title = {{VR} {Sickness} {Prediction} for {Navigation} in {Immersive} {Virtual} {Environments} using a {Deep} {Long} {Short} {Term} {Memory} {Model}},
	isbn = {978-1-7281-1377-7},
	url = {https://ieeexplore.ieee.org/document/8798213/},
	doi = {10.1109/VR.2019.8798213},
	language = {en},
	urldate = {2024-03-10},
	booktitle = {2019 {IEEE} {Conference} on {Virtual} {Reality} and {3D} {User} {Interfaces} ({VR})},
	publisher = {IEEE},
	author = {Wang, Yuyang and Chardonnet, Jean-Remy and Merienne, Frederic},
	month = mar,
	year = {2019},
	pages = {1874--1881},
}

@inproceedings{wang_speed_2018,
	address = {Maui, HI},
	title = {Speed {Profile} {Optimization} for {Enhanced} {Passenger} {Comfort}: {An} {Optimal} {Control} {Approach}},
	isbn = {978-1-7281-0321-1 978-1-7281-0323-5},
	shorttitle = {Speed {Profile} {Optimization} for {Enhanced} {Passenger} {Comfort}},
	url = {https://ieeexplore.ieee.org/document/8569420/},
	doi = {10.1109/ITSC.2018.8569420},
	abstract = {Autonomous vehicles are expected to start reaching the market within the next years. However in practical applications, navigation inside dynamic environments has to take many factors such as speed control, safety and comfort into consideration, which is more paramount for both passengers and pedestrians. In this paper, a novel speed profile planner based on an optimal control approach considering passenger comfort is proposed. The approach is accomplished by minimizing jerk under certain comfort constraints, which inherently gives a speed profile for the central nervous system to follow naturally. Imposed with the same conditions, the widely used Jerk Limitation method is interpreted as an equivalent of the minimum time control method, the latter being used to verify that our method can ensure better continuity and smoothness of the speed profiles. A validation test was specifically designed and performed in order to show the feasibility of our method.},
	language = {en},
	urldate = {2026-06-09},
	booktitle = {2018 21st {International} {Conference} on {Intelligent} {Transportation} {Systems} ({ITSC})},
	publisher = {IEEE},
	author = {Wang, Yuyang and Chardonnet, Jean-Remy and Merienne, Frederic},
	month = nov,
	year = {2018},
	pages = {723--728},
}

Autonomous vehicles are expected to start reaching the market within the next years. However in practical applications, navigation inside dynamic environments has to take many factors such as speed control, safety and comfort into consideration, which is more paramount for both passengers and pedestrians. In this paper, a novel speed profile planner based on an optimal control approach considering passenger comfort is proposed. The approach is accomplished by minimizing jerk under certain comfort constraints, which inherently gives a speed profile for the central nervous system to follow naturally. Imposed with the same conditions, the widely used Jerk Limitation method is interpreted as an equivalent of the minimum time control method, the latter being used to verify that our method can ensure better continuity and smoothness of the speed profiles. A validation test was specifically designed and performed in order to show the feasibility of our method.