AHCI RESEARCH GROUP
Publications
Papers published in international journals,
proceedings of conferences, workshops and books.
OUR RESEARCH
Scientific Publications
How to
Here you can find the complete list of our publications.
You can use the tag cloud to select only the papers dealing with specific research topics.
You can expand the Abstract, Links and BibTex record for each paper.
You can use the tag cloud to select only the papers dealing with specific research topics.
You can expand the Abstract, Links and BibTex record for each paper.
2025
Li, Z.; Zhang, H.; Peng, C.; Peiris, R.
Exploring Large Language Model-Driven Agents for Environment-Aware Spatial Interactions and Conversations in Virtual Reality Role-Play Scenarios Proceedings Article
In: Proc. - IEEE Conf. Virtual Real. 3D User Interfaces, VR, pp. 1–11, Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 979-833153645-9 (ISBN).
Abstract | Links | BibTeX | Tags: Chatbots, Computer simulation languages, Context- awareness, context-awareness, Digital elevation model, Generative AI, Human-AI Interaction, Language Model, Large language model, large language models, Model agents, Role-play simulation, role-play simulations, Role-plays, Spatial interaction, Virtual environments, Virtual Reality, Virtual-reality environment
@inproceedings{li_exploring_2025,
title = {Exploring Large Language Model-Driven Agents for Environment-Aware Spatial Interactions and Conversations in Virtual Reality Role-Play Scenarios},
author = {Z. Li and H. Zhang and C. Peng and R. Peiris},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-105002706893&doi=10.1109%2fVR59515.2025.00025&partnerID=40&md5=60f22109e054c9035a0c2210bb797039},
doi = {10.1109/VR59515.2025.00025},
isbn = {979-833153645-9 (ISBN)},
year = {2025},
date = {2025-01-01},
booktitle = {Proc. - IEEE Conf. Virtual Real. 3D User Interfaces, VR},
pages = {1–11},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Recent research has begun adopting Large Language Model (LLM) agents to enhance Virtual Reality (VR) interactions, creating immersive chatbot experiences. However, while current studies focus on generating dialogue from user speech inputs, their abilities to generate richer experiences based on the perception of LLM agents' VR environments and interaction cues remain unexplored. Hence, in this work, we propose an approach that enables LLM agents to perceive virtual environments and generate environment-aware interactions and conversations for an embodied human-AI interaction experience in VR environments. Here, we define a schema for describing VR environments and their interactions through text prompts. We evaluate the performance of our method through five role-play scenarios created using our approach in a study with 14 participants. The findings discuss the opportunities and challenges of our proposed approach for developing environment-aware LLM agents that facilitate spatial interactions and conversations within VR role-play scenarios. © 2025 IEEE.},
keywords = {Chatbots, Computer simulation languages, Context- awareness, context-awareness, Digital elevation model, Generative AI, Human-AI Interaction, Language Model, Large language model, large language models, Model agents, Role-play simulation, role-play simulations, Role-plays, Spatial interaction, Virtual environments, Virtual Reality, Virtual-reality environment},
pubstate = {published},
tppubtype = {inproceedings}
}
Recent research has begun adopting Large Language Model (LLM) agents to enhance Virtual Reality (VR) interactions, creating immersive chatbot experiences. However, while current studies focus on generating dialogue from user speech inputs, their abilities to generate richer experiences based on the perception of LLM agents' VR environments and interaction cues remain unexplored. Hence, in this work, we propose an approach that enables LLM agents to perceive virtual environments and generate environment-aware interactions and conversations for an embodied human-AI interaction experience in VR environments. Here, we define a schema for describing VR environments and their interactions through text prompts. We evaluate the performance of our method through five role-play scenarios created using our approach in a study with 14 participants. The findings discuss the opportunities and challenges of our proposed approach for developing environment-aware LLM agents that facilitate spatial interactions and conversations within VR role-play scenarios. © 2025 IEEE.
Kai, W. -H.; Xing, K. -X.
Video-driven musical composition using large language model with memory-augmented state space Journal Article
In: Visual Computer, vol. 41, no. 5, pp. 3345–3357, 2025, ISSN: 01782789 (ISSN).
Abstract | Links | BibTeX | Tags: 'current, Associative storage, Augmented Reality, Augmented state space, Computer simulation languages, Computer system recovery, Distributed computer systems, HTTP, Language Model, Large language model, Long-term video-to-music generation, Mamba, Memory architecture, Memory-augmented, Modeling languages, Music, Musical composition, Natural language processing systems, Object oriented programming, Performance, Problem oriented languages, State space, State-space
@article{kai_video-driven_2025,
title = {Video-driven musical composition using large language model with memory-augmented state space},
author = {W. -H. Kai and K. -X. Xing},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-105001073242&doi=10.1007%2fs00371-024-03606-w&partnerID=40&md5=7ea24f13614a9a24caf418c37a10bd8c},
doi = {10.1007/s00371-024-03606-w},
issn = {01782789 (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {Visual Computer},
volume = {41},
number = {5},
pages = {3345–3357},
abstract = {The current landscape of research leveraging large language models (LLMs) is experiencing a surge. Many works harness the powerful reasoning capabilities of these models to comprehend various modalities, such as text, speech, images, videos, etc. However, the research work on LLms for music inspiration is still in its infancy. To fill the gap in this field and break through the dilemma that LLMs can only understand short videos with limited frames, we propose a large language model with state space for long-term video-to-music generation. To capture long-range dependency and maintaining high performance, while further decrease the computing cost, our overall network includes the Enhanced Video Mamba, which incorporates continuous moving window partitioning and local feature augmentation, and a long-term memory bank that captures and aggregates historical video information to mitigate information loss in long sequences. This framework achieves both subquadratic-time computation and near-linear memory complexity, enabling effective long-term video-to-music generation. We conduct a thorough evaluation of our proposed framework. The experimental results demonstrate that our model achieves or surpasses the performance of the current state-of-the-art models. Our code released on https://github.com/kai211233/S2L2-V2M. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024.},
keywords = {'current, Associative storage, Augmented Reality, Augmented state space, Computer simulation languages, Computer system recovery, Distributed computer systems, HTTP, Language Model, Large language model, Long-term video-to-music generation, Mamba, Memory architecture, Memory-augmented, Modeling languages, Music, Musical composition, Natural language processing systems, Object oriented programming, Performance, Problem oriented languages, State space, State-space},
pubstate = {published},
tppubtype = {article}
}
The current landscape of research leveraging large language models (LLMs) is experiencing a surge. Many works harness the powerful reasoning capabilities of these models to comprehend various modalities, such as text, speech, images, videos, etc. However, the research work on LLms for music inspiration is still in its infancy. To fill the gap in this field and break through the dilemma that LLMs can only understand short videos with limited frames, we propose a large language model with state space for long-term video-to-music generation. To capture long-range dependency and maintaining high performance, while further decrease the computing cost, our overall network includes the Enhanced Video Mamba, which incorporates continuous moving window partitioning and local feature augmentation, and a long-term memory bank that captures and aggregates historical video information to mitigate information loss in long sequences. This framework achieves both subquadratic-time computation and near-linear memory complexity, enabling effective long-term video-to-music generation. We conduct a thorough evaluation of our proposed framework. The experimental results demonstrate that our model achieves or surpasses the performance of the current state-of-the-art models. Our code released on https://github.com/kai211233/S2L2-V2M. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024.
Chen, J.; Wu, X.; Lan, T.; Li, B.
LLMER: Crafting Interactive Extended Reality Worlds with JSON Data Generated by Large Language Models Journal Article
In: IEEE Transactions on Visualization and Computer Graphics, vol. 31, no. 5, pp. 2715–2724, 2025, ISSN: 10772626 (ISSN).
Abstract | Links | BibTeX | Tags: % reductions, 3D modeling, algorithm, Algorithms, Augmented Reality, Coding errors, Computer graphics, Computer interaction, computer interface, Computer simulation languages, Extended reality, generative artificial intelligence, human, Human users, human-computer interaction, Humans, Imaging, Immersive, Language, Language Model, Large language model, large language models, Metadata, Natural Language Processing, Natural language processing systems, Natural languages, procedures, Script generation, Spatio-temporal data, Three dimensional computer graphics, Three-Dimensional, three-dimensional imaging, User-Computer Interface, Virtual Reality
@article{chen_llmer_2025,
title = {LLMER: Crafting Interactive Extended Reality Worlds with JSON Data Generated by Large Language Models},
author = {J. Chen and X. Wu and T. Lan and B. Li},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-105003825793&doi=10.1109%2fTVCG.2025.3549549&partnerID=40&md5=da4681d0714548e3a7e0c8c3295d2348},
doi = {10.1109/TVCG.2025.3549549},
issn = {10772626 (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {IEEE Transactions on Visualization and Computer Graphics},
volume = {31},
number = {5},
pages = {2715–2724},
abstract = {The integration of Large Language Models (LLMs) like GPT-4 with Extended Reality (XR) technologies offers the potential to build truly immersive XR environments that interact with human users through natural language, e.g., generating and animating 3D scenes from audio inputs. However, the complexity of XR environments makes it difficult to accurately extract relevant contextual data and scene/object parameters from an overwhelming volume of XR artifacts. It leads to not only increased costs with pay-per-use models, but also elevated levels of generation errors. Moreover, existing approaches focusing on coding script generation are often prone to generation errors, resulting in flawed or invalid scripts, application crashes, and ultimately a degraded user experience. To overcome these challenges, we introduce LLMER, a novel framework that creates interactive XR worlds using JSON data generated by LLMs. Unlike prior approaches focusing on coding script generation, LLMER translates natural language inputs into JSON data, significantly reducing the likelihood of application crashes and processing latency. It employs a multi-stage strategy to supply only the essential contextual information adapted to the user's request and features multiple modules designed for various XR tasks. Our preliminary user study reveals the effectiveness of the proposed system, with over 80% reduction in consumed tokens and around 60% reduction in task completion time compared to state-of-the-art approaches. The analysis of users' feedback also illuminates a series of directions for further optimization. © 1995-2012 IEEE.},
keywords = {% reductions, 3D modeling, algorithm, Algorithms, Augmented Reality, Coding errors, Computer graphics, Computer interaction, computer interface, Computer simulation languages, Extended reality, generative artificial intelligence, human, Human users, human-computer interaction, Humans, Imaging, Immersive, Language, Language Model, Large language model, large language models, Metadata, Natural Language Processing, Natural language processing systems, Natural languages, procedures, Script generation, Spatio-temporal data, Three dimensional computer graphics, Three-Dimensional, three-dimensional imaging, User-Computer Interface, Virtual Reality},
pubstate = {published},
tppubtype = {article}
}
The integration of Large Language Models (LLMs) like GPT-4 with Extended Reality (XR) technologies offers the potential to build truly immersive XR environments that interact with human users through natural language, e.g., generating and animating 3D scenes from audio inputs. However, the complexity of XR environments makes it difficult to accurately extract relevant contextual data and scene/object parameters from an overwhelming volume of XR artifacts. It leads to not only increased costs with pay-per-use models, but also elevated levels of generation errors. Moreover, existing approaches focusing on coding script generation are often prone to generation errors, resulting in flawed or invalid scripts, application crashes, and ultimately a degraded user experience. To overcome these challenges, we introduce LLMER, a novel framework that creates interactive XR worlds using JSON data generated by LLMs. Unlike prior approaches focusing on coding script generation, LLMER translates natural language inputs into JSON data, significantly reducing the likelihood of application crashes and processing latency. It employs a multi-stage strategy to supply only the essential contextual information adapted to the user's request and features multiple modules designed for various XR tasks. Our preliminary user study reveals the effectiveness of the proposed system, with over 80% reduction in consumed tokens and around 60% reduction in task completion time compared to state-of-the-art approaches. The analysis of users' feedback also illuminates a series of directions for further optimization. © 1995-2012 IEEE.
Chen, J.; Grubert, J.; Kristensson, P. O.
Analyzing Multimodal Interaction Strategies for LLM-Assisted Manipulation of 3D Scenes Proceedings Article
In: Proc. - IEEE Conf. Virtual Real. 3D User Interfaces, VR, pp. 206–216, Institute of Electrical and Electronics Engineers Inc., 2025, ISBN: 979-833153645-9 (ISBN).
Abstract | Links | BibTeX | Tags: 3D modeling, 3D reconstruction, 3D scene editing, 3D scenes, Computer simulation languages, Editing systems, Immersive environment, Interaction pattern, Interaction strategy, Language Model, Large language model, large language models, Multimodal Interaction, Scene editing, Three dimensional computer graphics, Virtual environments, Virtual Reality
@inproceedings{chen_analyzing_2025,
title = {Analyzing Multimodal Interaction Strategies for LLM-Assisted Manipulation of 3D Scenes},
author = {J. Chen and J. Grubert and P. O. Kristensson},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-105002716635&doi=10.1109%2fVR59515.2025.00045&partnerID=40&md5=306aa7fbb3dad0aa9d43545f3c7eb9ea},
doi = {10.1109/VR59515.2025.00045},
isbn = {979-833153645-9 (ISBN)},
year = {2025},
date = {2025-01-01},
booktitle = {Proc. - IEEE Conf. Virtual Real. 3D User Interfaces, VR},
pages = {206–216},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {As more applications of large language models (LLMs) for 3D content in immersive environments emerge, it is crucial to study user behavior to identify interaction patterns and potential barriers to guide the future design of immersive content creation and editing systems which involve LLMs. In an empirical user study with 12 participants, we combine quantitative usage data with post-experience questionnaire feedback to reveal common interaction patterns and key barriers in LLM-assisted 3D scene editing systems. We identify opportunities for improving natural language interfaces in 3D design tools and propose design recommendations. Through an empirical study, we demonstrate that LLM-assisted interactive systems can be used productively in immersive environments. © 2025 IEEE.},
keywords = {3D modeling, 3D reconstruction, 3D scene editing, 3D scenes, Computer simulation languages, Editing systems, Immersive environment, Interaction pattern, Interaction strategy, Language Model, Large language model, large language models, Multimodal Interaction, Scene editing, Three dimensional computer graphics, Virtual environments, Virtual Reality},
pubstate = {published},
tppubtype = {inproceedings}
}
As more applications of large language models (LLMs) for 3D content in immersive environments emerge, it is crucial to study user behavior to identify interaction patterns and potential barriers to guide the future design of immersive content creation and editing systems which involve LLMs. In an empirical user study with 12 participants, we combine quantitative usage data with post-experience questionnaire feedback to reveal common interaction patterns and key barriers in LLM-assisted 3D scene editing systems. We identify opportunities for improving natural language interfaces in 3D design tools and propose design recommendations. Through an empirical study, we demonstrate that LLM-assisted interactive systems can be used productively in immersive environments. © 2025 IEEE.
Gaglio, G. F.; Vinanzi, S.; Cangelosi, A.; Chella, A.
Intention Reading Architecture for Virtual Agents Proceedings Article
In: O., Palinko; L., Bodenhagen; J.-J., Cabibihan; K., Fischer; S., Šabanović; K., Winkle; L., Behera; S.S., Ge; D., Chrysostomou; W., Jiang; H., He (Ed.): Lect. Notes Comput. Sci., pp. 488–497, Springer Science and Business Media Deutschland GmbH, 2025, ISBN: 03029743 (ISSN); 978-981963521-4 (ISBN).
Abstract | Links | BibTeX | Tags: Chatbots, Cognitive Architecture, Cognitive Architectures, Computer simulation languages, Intelligent virtual agents, Intention Reading, Intention readings, Language Model, Large language model, Metaverse, Metaverses, Physical robots, Video-games, Virtual agent, Virtual assistants, Virtual contexts, Virtual environments, Virtual machine
@inproceedings{gaglio_intention_2025,
title = {Intention Reading Architecture for Virtual Agents},
author = {G. F. Gaglio and S. Vinanzi and A. Cangelosi and A. Chella},
editor = {Palinko O. and Bodenhagen L. and Cabibihan J.-J. and Fischer K. and Šabanović S. and Winkle K. and Behera L. and Ge S.S. and Chrysostomou D. and Jiang W. and He H.},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-105002042645&doi=10.1007%2f978-981-96-3522-1_41&partnerID=40&md5=70ccc7039785bb4ca4d45752f1d3587f},
doi = {10.1007/978-981-96-3522-1_41},
isbn = {03029743 (ISSN); 978-981963521-4 (ISBN)},
year = {2025},
date = {2025-01-01},
booktitle = {Lect. Notes Comput. Sci.},
volume = {15561 LNAI},
pages = {488–497},
publisher = {Springer Science and Business Media Deutschland GmbH},
abstract = {This work presents the development of a virtual agent designed specifically for use in the Metaverse, video games, and other virtual environments, capable of performing intention reading on a human-controlled avatar through a cognitive architecture that endows it with contextual awareness. The paper explores the adaptation of a cognitive architecture, originally developed for physical robots, to a fully virtual context, where it is integrated with a Large Language Model to create highly communicative virtual assistants. Although this work primarily focuses on virtual applications, integrating cognitive architectures with LLMs marks a significant step toward creating collaborative artificial agents capable of providing meaningful support by deeply understanding context and user intentions in digital environments. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.},
keywords = {Chatbots, Cognitive Architecture, Cognitive Architectures, Computer simulation languages, Intelligent virtual agents, Intention Reading, Intention readings, Language Model, Large language model, Metaverse, Metaverses, Physical robots, Video-games, Virtual agent, Virtual assistants, Virtual contexts, Virtual environments, Virtual machine},
pubstate = {published},
tppubtype = {inproceedings}
}
This work presents the development of a virtual agent designed specifically for use in the Metaverse, video games, and other virtual environments, capable of performing intention reading on a human-controlled avatar through a cognitive architecture that endows it with contextual awareness. The paper explores the adaptation of a cognitive architecture, originally developed for physical robots, to a fully virtual context, where it is integrated with a Large Language Model to create highly communicative virtual assistants. Although this work primarily focuses on virtual applications, integrating cognitive architectures with LLMs marks a significant step toward creating collaborative artificial agents capable of providing meaningful support by deeply understanding context and user intentions in digital environments. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.
Xing, Y.; Ban, J.; Hubbard, T. D.; Villano, M.; Gómez-Zará, D.
Immersed in my Ideas: Using Virtual Reality and LLMs to Visualize Users’ Ideas and Thoughts Proceedings Article
In: Int Conf Intell User Interfaces Proc IUI, pp. 60–65, Association for Computing Machinery, 2025, ISBN: 979-840071409-2 (ISBN).
Abstract | Links | BibTeX | Tags: 3-D environments, 3D modeling, Computer simulation languages, Creativity, Idea Generation, Immersive, Interactive virtual reality, Language Model, Large language model, Multimodal Interaction, Reflection, Text Visualization, Think aloud, Virtual environments, Virtual Reality, Visualization
@inproceedings{xing_immersed_2025,
title = {Immersed in my Ideas: Using Virtual Reality and LLMs to Visualize Users’ Ideas and Thoughts},
author = {Y. Xing and J. Ban and T. D. Hubbard and M. Villano and D. Gómez-Zará},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-105001675169&doi=10.1145%2f3708557.3716330&partnerID=40&md5=20fb0623d2a1fff92282116b01fac4f3},
doi = {10.1145/3708557.3716330},
isbn = {979-840071409-2 (ISBN)},
year = {2025},
date = {2025-01-01},
booktitle = {Int Conf Intell User Interfaces Proc IUI},
pages = {60–65},
publisher = {Association for Computing Machinery},
abstract = {We introduce the Voice Interactive Virtual Reality Annotation (VIVRA), an application that employs Large Language Models to facilitate brainstorming and idea exploration in an immersive 3D environment. As users think aloud to brainstorm and ideate, the application automatically detects, summarizes, suggests, and connects their ideas in real time. The experience brings participants into a room where their ideas emerge as interactive objects that embody the topics detected from their ideas. We evaluated the effectiveness of VIVRA in an exploratory study with 29 participants, followed by a user study with 10 participants comparing the application with other visualizations. Our results show that VIVRA helped participants reflect and think more about their ideas, serving as a valuable tool for personal exploration. We discuss the potential benefits and applications, highlighting the benefits of combining immersive 3D spaces and LLMs to explore, learn, and reflect on ideas. © 2025 Copyright held by the owner/author(s).},
keywords = {3-D environments, 3D modeling, Computer simulation languages, Creativity, Idea Generation, Immersive, Interactive virtual reality, Language Model, Large language model, Multimodal Interaction, Reflection, Text Visualization, Think aloud, Virtual environments, Virtual Reality, Visualization},
pubstate = {published},
tppubtype = {inproceedings}
}
We introduce the Voice Interactive Virtual Reality Annotation (VIVRA), an application that employs Large Language Models to facilitate brainstorming and idea exploration in an immersive 3D environment. As users think aloud to brainstorm and ideate, the application automatically detects, summarizes, suggests, and connects their ideas in real time. The experience brings participants into a room where their ideas emerge as interactive objects that embody the topics detected from their ideas. We evaluated the effectiveness of VIVRA in an exploratory study with 29 participants, followed by a user study with 10 participants comparing the application with other visualizations. Our results show that VIVRA helped participants reflect and think more about their ideas, serving as a valuable tool for personal exploration. We discuss the potential benefits and applications, highlighting the benefits of combining immersive 3D spaces and LLMs to explore, learn, and reflect on ideas. © 2025 Copyright held by the owner/author(s).
Kurai, R.; Hiraki, T.; Hiroi, Y.; Hirao, Y.; Perusquia-Hernandez, M.; Uchiyama, H.; Kiyokawa, K.
MagicItem: Dynamic Behavior Design of Virtual Objects With Large Language Models in a Commercial Metaverse Platform Journal Article
In: IEEE Access, vol. 13, pp. 19132–19143, 2025, ISSN: 21693536 (ISSN).
Abstract | Links | BibTeX | Tags: Behavior design, Code programming, Computer simulation languages, Dynamic behaviors, Language Model, Large-language model, Low-code programming, Metaverse platform, Metaverses, Virtual addresses, Virtual environments, Virtual objects, Virtual Reality, Virtual-reality environment
@article{kurai_magicitem_2025,
title = {MagicItem: Dynamic Behavior Design of Virtual Objects With Large Language Models in a Commercial Metaverse Platform},
author = {R. Kurai and T. Hiraki and Y. Hiroi and Y. Hirao and M. Perusquia-Hernandez and H. Uchiyama and K. Kiyokawa},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85216011970&doi=10.1109%2fACCESS.2025.3530439&partnerID=40&md5=7a33b9618af8b4ab79b43fb3bd4317cf},
doi = {10.1109/ACCESS.2025.3530439},
issn = {21693536 (ISSN)},
year = {2025},
date = {2025-01-01},
journal = {IEEE Access},
volume = {13},
pages = {19132–19143},
abstract = {To create rich experiences in virtual reality (VR) environments, it is essential to define the behavior of virtual objects through programming. However, programming in 3D spaces requires a wide range of background knowledge and programming skills. Although Large Language Models (LLMs) have provided programming support, they are still primarily aimed at programmers. In metaverse platforms, where many users inhabit VR spaces, most users are unfamiliar with programming, making it difficult for them to modify the behavior of objects in the VR environment easily. Existing LLM-based script generation methods for VR spaces require multiple lengthy iterations to implement the desired behaviors and are difficult to integrate into the operation of metaverse platforms. To address this issue, we propose a tool that generates behaviors for objects in VR spaces from natural language within Cluster, a metaverse platform with a large user base. By integrating LLMs with the Cluster Script provided by this platform, we enable users with limited programming experience to define object behaviors within the platform freely. We have also integrated our tool into a commercial metaverse platform and are conducting online experiments with 63 general users of the platform. The experiments show that even users with no programming background can successfully generate behaviors for objects in VR spaces, resulting in a highly satisfying system. Our research contributes to democratizing VR content creation by enabling non-programmers to design dynamic behaviors for virtual objects in metaverse platforms. © 2013 IEEE.},
keywords = {Behavior design, Code programming, Computer simulation languages, Dynamic behaviors, Language Model, Large-language model, Low-code programming, Metaverse platform, Metaverses, Virtual addresses, Virtual environments, Virtual objects, Virtual Reality, Virtual-reality environment},
pubstate = {published},
tppubtype = {article}
}
To create rich experiences in virtual reality (VR) environments, it is essential to define the behavior of virtual objects through programming. However, programming in 3D spaces requires a wide range of background knowledge and programming skills. Although Large Language Models (LLMs) have provided programming support, they are still primarily aimed at programmers. In metaverse platforms, where many users inhabit VR spaces, most users are unfamiliar with programming, making it difficult for them to modify the behavior of objects in the VR environment easily. Existing LLM-based script generation methods for VR spaces require multiple lengthy iterations to implement the desired behaviors and are difficult to integrate into the operation of metaverse platforms. To address this issue, we propose a tool that generates behaviors for objects in VR spaces from natural language within Cluster, a metaverse platform with a large user base. By integrating LLMs with the Cluster Script provided by this platform, we enable users with limited programming experience to define object behaviors within the platform freely. We have also integrated our tool into a commercial metaverse platform and are conducting online experiments with 63 general users of the platform. The experiments show that even users with no programming background can successfully generate behaviors for objects in VR spaces, resulting in a highly satisfying system. Our research contributes to democratizing VR content creation by enabling non-programmers to design dynamic behaviors for virtual objects in metaverse platforms. © 2013 IEEE.
2024
Lee, L. -K.; Chan, E. H.; Tong, K. K. -L.; Wong, N. K. -H.; Wu, B. S. -Y.; Fung, Y. -C.; Fong, E. K. S.; Hou, U. Leong; Wu, N. -I.
Utilizing Virtual Reality and Generative AI Chatbot for Job Interview Simulations Proceedings Article
In: K.T., Chui; Y.K., Hui; D., Yang; L.-K., Lee; L.-P., Wong; B.L., Reynolds (Ed.): Proc. - Int. Symp. Educ. Technol., ISET, pp. 209–212, Institute of Electrical and Electronics Engineers Inc., 2024, ISBN: 979-835036141-4 (ISBN).
Abstract | Links | BibTeX | Tags: chatbot, Chatbots, Computer interaction, Computer simulation languages, Generative adversarial networks, Generative AI, Hong-kong, Human computer interaction, ITS applications, Job interview simulation, Job interviews, Performance, Science graduates, User friendliness, Virtual environments, Virtual Reality
@inproceedings{lee_utilizing_2024,
title = {Utilizing Virtual Reality and Generative AI Chatbot for Job Interview Simulations},
author = {L. -K. Lee and E. H. Chan and K. K. -L. Tong and N. K. -H. Wong and B. S. -Y. Wu and Y. -C. Fung and E. K. S. Fong and U. Leong Hou and N. -I. Wu},
editor = {Chui K.T. and Hui Y.K. and Yang D. and Lee L.-K. and Wong L.-P. and Reynolds B.L.},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85206582338&doi=10.1109%2fISET61814.2024.00048&partnerID=40&md5=c6986c0697792254e167e143b75f14c6},
doi = {10.1109/ISET61814.2024.00048},
isbn = {979-835036141-4 (ISBN)},
year = {2024},
date = {2024-01-01},
booktitle = {Proc. - Int. Symp. Educ. Technol., ISET},
pages = {209–212},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {Stress and anxiety experienced by interviewees, particularly fresh graduates, would significantly impact their performance in job interviews. Due to the increased affordability and user-friendliness of virtual reality (VR), VR has seen a surge in its application within the educational sector. This paper presents the design and implementation of a job interview simulation system, leveraging VR and a generative AI chatbot to provide an immersive environment for computer science graduates in Hong Kong. The system aims to help graduates practice and familiarize themselves with various real-world scenarios of a job interview in English, Mandarin, and Cantonese, tailored to the unique language requirements of Hong Kong's professional environment. The system comprises three core modules: a mock question and answer reading module, an AI speech analysis module, and a virtual interview module facilitated by the generative AI chatbot, ChatGPT. We anticipate that the proposed simulator will provide valuable insights to education practitioners on utilizing VR and generative AI for job interview training, extending beyond computer science graduates. © 2024 IEEE.},
keywords = {chatbot, Chatbots, Computer interaction, Computer simulation languages, Generative adversarial networks, Generative AI, Hong-kong, Human computer interaction, ITS applications, Job interview simulation, Job interviews, Performance, Science graduates, User friendliness, Virtual environments, Virtual Reality},
pubstate = {published},
tppubtype = {inproceedings}
}
Stress and anxiety experienced by interviewees, particularly fresh graduates, would significantly impact their performance in job interviews. Due to the increased affordability and user-friendliness of virtual reality (VR), VR has seen a surge in its application within the educational sector. This paper presents the design and implementation of a job interview simulation system, leveraging VR and a generative AI chatbot to provide an immersive environment for computer science graduates in Hong Kong. The system aims to help graduates practice and familiarize themselves with various real-world scenarios of a job interview in English, Mandarin, and Cantonese, tailored to the unique language requirements of Hong Kong's professional environment. The system comprises three core modules: a mock question and answer reading module, an AI speech analysis module, and a virtual interview module facilitated by the generative AI chatbot, ChatGPT. We anticipate that the proposed simulator will provide valuable insights to education practitioners on utilizing VR and generative AI for job interview training, extending beyond computer science graduates. © 2024 IEEE.
Jayaraman, S.; Bhavya, R.; Srihari, V.; Rajam, V. Mary Anita
TexAVi: Generating Stereoscopic VR Video Clips from Text Descriptions Proceedings Article
In: IEEE Int. Conf. Comput. Vis. Mach. Intell., CVMI, Institute of Electrical and Electronics Engineers Inc., 2024, ISBN: 979-835037687-6 (ISBN).
Abstract | Links | BibTeX | Tags: Adversarial networks, Computer simulation languages, Deep learning, Depth Estimation, Depth perception, Diffusion Model, diffusion models, Digital elevation model, Generative adversarial networks, Generative model, Generative systems, Language Model, Motion capture, Stereo image processing, Text-to-image, Training data, Video analysis, Video-clips, Virtual environments, Virtual Reality
@inproceedings{jayaraman_texavi_2024,
title = {TexAVi: Generating Stereoscopic VR Video Clips from Text Descriptions},
author = {S. Jayaraman and R. Bhavya and V. Srihari and V. Mary Anita Rajam},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85215265234&doi=10.1109%2fCVMI61877.2024.10782691&partnerID=40&md5=8e20576af67b917ecfad83873a87ef29},
doi = {10.1109/CVMI61877.2024.10782691},
isbn = {979-835037687-6 (ISBN)},
year = {2024},
date = {2024-01-01},
booktitle = {IEEE Int. Conf. Comput. Vis. Mach. Intell., CVMI},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {While generative models such as text-to-image, large language models and text-to-video have seen significant progress, the extension to text-to-virtual-reality remains largely unexplored, due to a deficit in training data and the complexity of achieving realistic depth and motion in virtual environments. This paper proposes an approach to coalesce existing generative systems to form a stereoscopic virtual reality video from text. Carried out in three main stages, we start with a base text-to-image model that captures context from an input text. We then employ Stable Diffusion on the rudimentary image produced, to generate frames with enhanced realism and overall quality. These frames are processed with depth estimation algorithms to create left-eye and right-eye views, which are stitched side-by-side to create an immersive viewing experience. Such systems would be highly beneficial in virtual reality production, since filming and scene building often require extensive hours of work and post-production effort. We utilize image evaluation techniques, specifically Fréchet Inception Distance and CLIP Score, to assess the visual quality of frames produced for the video. These quantitative measures establish the proficiency of the proposed method. Our work highlights the exciting possibilities of using natural language-driven graphics in fields like virtual reality simulations. © 2024 IEEE.},
keywords = {Adversarial networks, Computer simulation languages, Deep learning, Depth Estimation, Depth perception, Diffusion Model, diffusion models, Digital elevation model, Generative adversarial networks, Generative model, Generative systems, Language Model, Motion capture, Stereo image processing, Text-to-image, Training data, Video analysis, Video-clips, Virtual environments, Virtual Reality},
pubstate = {published},
tppubtype = {inproceedings}
}
While generative models such as text-to-image, large language models and text-to-video have seen significant progress, the extension to text-to-virtual-reality remains largely unexplored, due to a deficit in training data and the complexity of achieving realistic depth and motion in virtual environments. This paper proposes an approach to coalesce existing generative systems to form a stereoscopic virtual reality video from text. Carried out in three main stages, we start with a base text-to-image model that captures context from an input text. We then employ Stable Diffusion on the rudimentary image produced, to generate frames with enhanced realism and overall quality. These frames are processed with depth estimation algorithms to create left-eye and right-eye views, which are stitched side-by-side to create an immersive viewing experience. Such systems would be highly beneficial in virtual reality production, since filming and scene building often require extensive hours of work and post-production effort. We utilize image evaluation techniques, specifically Fréchet Inception Distance and CLIP Score, to assess the visual quality of frames produced for the video. These quantitative measures establish the proficiency of the proposed method. Our work highlights the exciting possibilities of using natural language-driven graphics in fields like virtual reality simulations. © 2024 IEEE.
Geetha, S.; Aditya, G.; Reddy, M. Chetan; Nischith, G.
Human Interaction in Virtual and Mixed Reality Through Hand Tracking Proceedings Article
In: Proc. CONECCT - IEEE Int. Conf. Electron., Comput. Commun. Technol., Institute of Electrical and Electronics Engineers Inc., 2024, ISBN: 979-835038592-2 (ISBN).
Abstract | Links | BibTeX | Tags: Computer interaction, Computer simulation languages, Daily lives, Digital elevation model, Hand gesture, hand tracking, Hand-tracking, human-computer interaction, Humaninteraction, Interaction dynamics, Mixed reality, Unity, User friendly interface, User interfaces, Virtual environments, Virtual Reality, Virtual spaces
@inproceedings{geetha_human_2024,
title = {Human Interaction in Virtual and Mixed Reality Through Hand Tracking},
author = {S. Geetha and G. Aditya and M. Chetan Reddy and G. Nischith},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85205768661&doi=10.1109%2fCONECCT62155.2024.10677239&partnerID=40&md5=173e590ca9a1e30b760d05af562f311a},
doi = {10.1109/CONECCT62155.2024.10677239},
isbn = {979-835038592-2 (ISBN)},
year = {2024},
date = {2024-01-01},
booktitle = {Proc. CONECCT - IEEE Int. Conf. Electron., Comput. Commun. Technol.},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
abstract = {This paper explores the potential and possibilities of hand tracking in virtual reality (VR) and mixed reality (MR), focusing on its role in human interaction dynamics. An application was designed in Unity leveraging the XR Interaction toolkit, within which various items across three important domains: daily life, education, and recreation, were crafted to demonstrate the versatility of hand tracking along with hand gesture-based shortcuts for interaction. Integration of elements in MR ensures that users can seamlessly enjoy virtual experiences while remaining connected to their physical surroundings. Precise hand tracking enables effortless interaction with the virtual space, enhancing presence and control with a user-friendly interface. Additionally, the paper explores the effectiveness of integrating hand tracking into education and training scenarios. A computer assembly simulation was created to demonstrate this, featuring component inspection and zoom capabilities along with a large language model (LLM) integrated with hand gestures to provide for interaction capabilities. © 2024 IEEE.},
keywords = {Computer interaction, Computer simulation languages, Daily lives, Digital elevation model, Hand gesture, hand tracking, Hand-tracking, human-computer interaction, Humaninteraction, Interaction dynamics, Mixed reality, Unity, User friendly interface, User interfaces, Virtual environments, Virtual Reality, Virtual spaces},
pubstate = {published},
tppubtype = {inproceedings}
}
This paper explores the potential and possibilities of hand tracking in virtual reality (VR) and mixed reality (MR), focusing on its role in human interaction dynamics. An application was designed in Unity leveraging the XR Interaction toolkit, within which various items across three important domains: daily life, education, and recreation, were crafted to demonstrate the versatility of hand tracking along with hand gesture-based shortcuts for interaction. Integration of elements in MR ensures that users can seamlessly enjoy virtual experiences while remaining connected to their physical surroundings. Precise hand tracking enables effortless interaction with the virtual space, enhancing presence and control with a user-friendly interface. Additionally, the paper explores the effectiveness of integrating hand tracking into education and training scenarios. A computer assembly simulation was created to demonstrate this, featuring component inspection and zoom capabilities along with a large language model (LLM) integrated with hand gestures to provide for interaction capabilities. © 2024 IEEE.
Jiang, H.; Song, L.; Weng, D.; Sun, Z.; Li, H.; Dongye, X.; Zhang, Z.
In Situ 3D Scene Synthesis for Ubiquitous Embodied Interfaces Proceedings Article
In: MM - Proc. ACM Int. Conf. Multimed., pp. 3666–3675, Association for Computing Machinery, Inc, 2024, ISBN: 979-840070686-8 (ISBN).
Abstract | Links | BibTeX | Tags: 3D modeling, 3D scenes, affordance, Affordances, Chatbots, Computer simulation languages, Digital elevation model, Embodied interfaces, Language Model, Large language model, Physical environments, Scene synthesis, Synthesised, Three dimensional computer graphics, user demand, User demands, Virtual environments, Virtual Reality, Virtual scenes
@inproceedings{jiang_situ_2024,
title = {In Situ 3D Scene Synthesis for Ubiquitous Embodied Interfaces},
author = {H. Jiang and L. Song and D. Weng and Z. Sun and H. Li and X. Dongye and Z. Zhang},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85209812307&doi=10.1145%2f3664647.3681616&partnerID=40&md5=e58acd404c8785868c69a4647cecacb2},
doi = {10.1145/3664647.3681616},
isbn = {979-840070686-8 (ISBN)},
year = {2024},
date = {2024-01-01},
booktitle = {MM - Proc. ACM Int. Conf. Multimed.},
pages = {3666–3675},
publisher = {Association for Computing Machinery, Inc},
abstract = {Virtual reality enables us to access and interact with immersive virtual environments anytime and anywhere in various fields such as entertainment, training, and education. However, users immersed in virtual scenes remain physically connected to their real-world surroundings, which can pose safety and immersion challenges. Although virtual scene synthesis has attracted widespread attention, many popular methods are limited to generating purely virtual scenes independent of physical environments or simply mapping physical objects as obstacles. To this end, we propose a scene agent that synthesizes situated 3D virtual scenes as a kind of ubiquitous embodied interface in VR for users. The scene agent synthesizes scenes by perceiving the user's physical environment as well as inferring the user's demands. The synthesized scenes maintain the affordances of the physical environment, enabling immersive users to interact with the physical environment and improving the user's sense of security. Meanwhile, the synthesized scenes maintain the style described by the user, improving the user's immersion. The comparison results show that the proposed scene agent can synthesize virtual scenes with better affordance maintenance, scene diversity, style maintenance, and 3D intersection over union compared to baselines. To the best of our knowledge, this is the first work that achieves in situ scene synthesis with virtual-real affordance consistency and user demand. © 2024 ACM.},
keywords = {3D modeling, 3D scenes, affordance, Affordances, Chatbots, Computer simulation languages, Digital elevation model, Embodied interfaces, Language Model, Large language model, Physical environments, Scene synthesis, Synthesised, Three dimensional computer graphics, user demand, User demands, Virtual environments, Virtual Reality, Virtual scenes},
pubstate = {published},
tppubtype = {inproceedings}
}
Virtual reality enables us to access and interact with immersive virtual environments anytime and anywhere in various fields such as entertainment, training, and education. However, users immersed in virtual scenes remain physically connected to their real-world surroundings, which can pose safety and immersion challenges. Although virtual scene synthesis has attracted widespread attention, many popular methods are limited to generating purely virtual scenes independent of physical environments or simply mapping physical objects as obstacles. To this end, we propose a scene agent that synthesizes situated 3D virtual scenes as a kind of ubiquitous embodied interface in VR for users. The scene agent synthesizes scenes by perceiving the user's physical environment as well as inferring the user's demands. The synthesized scenes maintain the affordances of the physical environment, enabling immersive users to interact with the physical environment and improving the user's sense of security. Meanwhile, the synthesized scenes maintain the style described by the user, improving the user's immersion. The comparison results show that the proposed scene agent can synthesize virtual scenes with better affordance maintenance, scene diversity, style maintenance, and 3D intersection over union compared to baselines. To the best of our knowledge, this is the first work that achieves in situ scene synthesis with virtual-real affordance consistency and user demand. © 2024 ACM.
Christiansen, F. R.; Hollensberg, L. Nø.; Jensen, N. B.; Julsgaard, K.; Jespersen, K. N.; Nikolov, I.
Exploring Presence in Interactions with LLM-Driven NPCs: A Comparative Study of Speech Recognition and Dialogue Options Proceedings Article
In: S.N., Spencer (Ed.): Proc. ACM Symp. Virtual Reality Softw. Technol. VRST, Association for Computing Machinery, 2024, ISBN: 979-840070535-9 (ISBN).
Abstract | Links | BibTeX | Tags: Comparatives studies, Computer simulation languages, Economic and social effects, Immersive System, Immersive systems, Language Model, Large language model, Large language models (LLM), Model-driven, Modern technologies, Non-playable character, NPC, Presence, Social Actors, Speech enhancement, Speech recognition, Text to speech, Virtual environments, Virtual Reality, VR
@inproceedings{christiansen_exploring_2024,
title = {Exploring Presence in Interactions with LLM-Driven NPCs: A Comparative Study of Speech Recognition and Dialogue Options},
author = {F. R. Christiansen and L. Nø. Hollensberg and N. B. Jensen and K. Julsgaard and K. N. Jespersen and I. Nikolov},
editor = {Spencer S.N.},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85212512351&doi=10.1145%2f3641825.3687716&partnerID=40&md5=56ec6982b399fd97196ea73e7c659c31},
doi = {10.1145/3641825.3687716},
isbn = {979-840070535-9 (ISBN)},
year = {2024},
date = {2024-01-01},
booktitle = {Proc. ACM Symp. Virtual Reality Softw. Technol. VRST},
publisher = {Association for Computing Machinery},
abstract = {Combining modern technologies like large-language models (LLMs), speech-to-text, and text-to-speech can enhance immersion in virtual reality (VR) environments. However, challenges exist in effectively implementing LLMs and educating users. This paper explores implementing LLM-powered virtual social actors and facilitating user communication. We developed a murder mystery game where users interact with LLM-based non-playable characters (NPCs) through interrogation, clue-gathering, and exploration. Two versions were tested: one using speech recognition and another with traditional dialog boxes. While both provided similar social presence, users felt more immersed with speech recognition but found it overwhelming, while the dialog version was more challenging. Slow NPC response times were a source of frustration, highlighting the need for faster generation or better masking for a seamless experience. © 2024 Owner/Author.},
keywords = {Comparatives studies, Computer simulation languages, Economic and social effects, Immersive System, Immersive systems, Language Model, Large language model, Large language models (LLM), Model-driven, Modern technologies, Non-playable character, NPC, Presence, Social Actors, Speech enhancement, Speech recognition, Text to speech, Virtual environments, Virtual Reality, VR},
pubstate = {published},
tppubtype = {inproceedings}
}
Combining modern technologies like large-language models (LLMs), speech-to-text, and text-to-speech can enhance immersion in virtual reality (VR) environments. However, challenges exist in effectively implementing LLMs and educating users. This paper explores implementing LLM-powered virtual social actors and facilitating user communication. We developed a murder mystery game where users interact with LLM-based non-playable characters (NPCs) through interrogation, clue-gathering, and exploration. Two versions were tested: one using speech recognition and another with traditional dialog boxes. While both provided similar social presence, users felt more immersed with speech recognition but found it overwhelming, while the dialog version was more challenging. Slow NPC response times were a source of frustration, highlighting the need for faster generation or better masking for a seamless experience. © 2024 Owner/Author.
Tang, Y.; Situ, J.; Huang, Y.
Beyond User Experience: Technical and Contextual Metrics for Large Language Models in Extended Reality Proceedings Article
In: UbiComp Companion - Companion ACM Int. Jt. Conf. Pervasive Ubiquitous Comput., pp. 640–643, Association for Computing Machinery, Inc, 2024, ISBN: 979-840071058-2 (ISBN).
Abstract | Links | BibTeX | Tags: Augmented Reality, Computer simulation languages, Evaluation Metrics, Extended reality, Language Model, Large language model, large language models, Mixed reality, Modeling performance, Natural language processing systems, Physical world, Spatial computing, spatial data, user experience, Users' experiences, Virtual environments, Virtual Reality
@inproceedings{tang_beyond_2024,
title = {Beyond User Experience: Technical and Contextual Metrics for Large Language Models in Extended Reality},
author = {Y. Tang and J. Situ and Y. Huang},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85206203437&doi=10.1145%2f3675094.3678995&partnerID=40&md5=3fb337872b483a163bfbea038f1baffe},
doi = {10.1145/3675094.3678995},
isbn = {979-840071058-2 (ISBN)},
year = {2024},
date = {2024-01-01},
booktitle = {UbiComp Companion - Companion ACM Int. Jt. Conf. Pervasive Ubiquitous Comput.},
pages = {640–643},
publisher = {Association for Computing Machinery, Inc},
abstract = {Spatial Computing involves interacting with the physical world through spatial data manipulation, closely linked with Extended Reality (XR), which includes Virtual Reality (VR), Augmented Reality (AR), and Mixed Reality (MR). Large Language Models (LLMs) significantly enhance XR applications by improving user interactions through natural language understanding and content generation. Typical evaluations of these applications focus on user experience (UX) metrics, such as task performance, user satisfaction, and psychological assessments, but often neglect the technical performance of the LLMs themselves. This paper identifies significant gaps in current evaluation practices for LLMs within XR environments, attributing them to the novelty of the field, the complexity of spatial contexts, and the multimodal nature of interactions in XR. To address these gaps, the paper proposes specific metrics tailored to evaluate LLM performance in XR contexts, including spatial contextual awareness, coherence, proactivity, multimodal integration, hallucination, and question-answering accuracy. These proposed metrics aim to complement existing UX evaluations, providing a comprehensive assessment framework that captures both the technical and user-centric aspects of LLM performance in XR applications. The conclusion underscores the necessity for a dual-focused approach that combines technical and UX metrics to ensure effective and user-friendly LLM-integrated XR systems. © 2024 Copyright held by the owner/author(s).},
keywords = {Augmented Reality, Computer simulation languages, Evaluation Metrics, Extended reality, Language Model, Large language model, large language models, Mixed reality, Modeling performance, Natural language processing systems, Physical world, Spatial computing, spatial data, user experience, Users' experiences, Virtual environments, Virtual Reality},
pubstate = {published},
tppubtype = {inproceedings}
}
Spatial Computing involves interacting with the physical world through spatial data manipulation, closely linked with Extended Reality (XR), which includes Virtual Reality (VR), Augmented Reality (AR), and Mixed Reality (MR). Large Language Models (LLMs) significantly enhance XR applications by improving user interactions through natural language understanding and content generation. Typical evaluations of these applications focus on user experience (UX) metrics, such as task performance, user satisfaction, and psychological assessments, but often neglect the technical performance of the LLMs themselves. This paper identifies significant gaps in current evaluation practices for LLMs within XR environments, attributing them to the novelty of the field, the complexity of spatial contexts, and the multimodal nature of interactions in XR. To address these gaps, the paper proposes specific metrics tailored to evaluate LLM performance in XR contexts, including spatial contextual awareness, coherence, proactivity, multimodal integration, hallucination, and question-answering accuracy. These proposed metrics aim to complement existing UX evaluations, providing a comprehensive assessment framework that captures both the technical and user-centric aspects of LLM performance in XR applications. The conclusion underscores the necessity for a dual-focused approach that combines technical and UX metrics to ensure effective and user-friendly LLM-integrated XR systems. © 2024 Copyright held by the owner/author(s).
de Oliveira, E. A. Masasi; Silva, D. F. C.; Filho, A. R. G.
Improving VR Accessibility Through Automatic 360 Scene Description Using Multimodal Large Language Models Proceedings Article
In: ACM Int. Conf. Proc. Ser., pp. 289–293, Association for Computing Machinery, 2024, ISBN: 979-840070979-1 (ISBN).
Abstract | Links | BibTeX | Tags: 3D Scene, 3D scenes, Accessibility, Computer simulation languages, Descriptive information, Digital elevation model, Immersive, Language Model, Multi-modal, Multimodal large language model, Multimodal Large Language Models (MLLMs), Scene description, Virtual environments, Virtual Reality, Virtual Reality (VR), Virtual reality technology
@inproceedings{masasi_de_oliveira_improving_2024,
title = {Improving VR Accessibility Through Automatic 360 Scene Description Using Multimodal Large Language Models},
author = {E. A. Masasi de Oliveira and D. F. C. Silva and A. R. G. Filho},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85206580797&doi=10.1145%2f3691573.3691619&partnerID=40&md5=6e80800fce0e6b56679fbcbe982bcfa7},
doi = {10.1145/3691573.3691619},
isbn = {979-840070979-1 (ISBN)},
year = {2024},
date = {2024-01-01},
booktitle = {ACM Int. Conf. Proc. Ser.},
pages = {289–293},
publisher = {Association for Computing Machinery},
abstract = {Advancements in Virtual Reality (VR) technology hold immense promise for enriching immersive experiences. Despite the advancements in VR technology, there remains a significant gap in addressing accessibility concerns, particularly in automatically providing descriptive information for VR scenes. This paper combines the potential of leveraging Multimodal Large Language Models (MLLMs) to automatically generate text descriptions for 360 VR scenes according to Speech-to-Text (STT) prompts. As a case study, we conduct experiments on educational settings in VR museums, improving dynamic experiences across various contexts. Despite minor challenges in adapting MLLMs to VR Scenes, the experiments demonstrate that they can generate descriptions with high quality. Our findings provide insights for enhancing VR experiences and ensuring accessibility to individuals with disabilities or diverse needs. © 2024 Copyright held by the owner/author(s).},
keywords = {3D Scene, 3D scenes, Accessibility, Computer simulation languages, Descriptive information, Digital elevation model, Immersive, Language Model, Multi-modal, Multimodal large language model, Multimodal Large Language Models (MLLMs), Scene description, Virtual environments, Virtual Reality, Virtual Reality (VR), Virtual reality technology},
pubstate = {published},
tppubtype = {inproceedings}
}
Advancements in Virtual Reality (VR) technology hold immense promise for enriching immersive experiences. Despite the advancements in VR technology, there remains a significant gap in addressing accessibility concerns, particularly in automatically providing descriptive information for VR scenes. This paper combines the potential of leveraging Multimodal Large Language Models (MLLMs) to automatically generate text descriptions for 360 VR scenes according to Speech-to-Text (STT) prompts. As a case study, we conduct experiments on educational settings in VR museums, improving dynamic experiences across various contexts. Despite minor challenges in adapting MLLMs to VR Scenes, the experiments demonstrate that they can generate descriptions with high quality. Our findings provide insights for enhancing VR experiences and ensuring accessibility to individuals with disabilities or diverse needs. © 2024 Copyright held by the owner/author(s).
Kang, Z.; Liu, Y.; Zheng, J.; Sun, Z.
Revealing the Difficulty in Jailbreak Defense on Language Models for Metaverse Proceedings Article
In: Q., Gong; X., He (Ed.): SocialMeta - Proc. Int. Workshop Soc. Metaverse Comput., Sens. Netw., Part: ACM SenSys, pp. 31–37, Association for Computing Machinery, Inc, 2024, ISBN: 979-840071299-9 (ISBN).
Abstract | Links | BibTeX | Tags: % reductions, Attack strategies, Computer simulation languages, Defense, Digital elevation model, Guard rails, Jailbreak, Language Model, Large language model, Metaverse Security, Metaverses, Natural languages, Performance, Virtual Reality
@inproceedings{kang_revealing_2024,
title = {Revealing the Difficulty in Jailbreak Defense on Language Models for Metaverse},
author = {Z. Kang and Y. Liu and J. Zheng and Z. Sun},
editor = {Gong Q. and He X.},
url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85212189363&doi=10.1145%2f3698387.3699998&partnerID=40&md5=673326728c3db35ffbbaf807eb7f003c},
doi = {10.1145/3698387.3699998},
isbn = {979-840071299-9 (ISBN)},
year = {2024},
date = {2024-01-01},
booktitle = {SocialMeta - Proc. Int. Workshop Soc. Metaverse Comput., Sens. Netw., Part: ACM SenSys},
pages = {31–37},
publisher = {Association for Computing Machinery, Inc},
abstract = {Large language models (LLMs) have demonstrated exceptional capabilities in natural language processing tasks, fueling innovations in emerging areas such as the metaverse. These models enable dynamic virtual communities, enhancing user interactions and revolutionizing industries. However, their increasing deployment exposes vulnerabilities to jailbreak attacks, where adversaries can manipulate LLM-driven systems to generate harmful content. While various defense mechanisms have been proposed, their efficacy against diverse jailbreak techniques remains unclear. This paper addresses this gap by evaluating the performance of three popular defense methods (Backtranslation, Self-reminder, and Paraphrase) against different jailbreak attack strategies (GCG, BEAST, and Deepinception), while also utilizing three distinct models. Our findings reveal that while defenses are highly effective against optimization-based jailbreak attacks and reduce the attack success rate by 79% on average, they struggle in defending against attacks that alter attack motivations. Additionally, methods relying on self-reminding perform better when integrated with models featuring robust safety guardrails. For instance, Llama2-7b shows a 100% reduction in Attack Success Rate, while Vicuna-7b and Mistral-7b, lacking safety alignment, exhibit a lower average reduction of 65.8%. This study highlights the challenges in developing universal defense solutions for securing LLMs in dynamic environments like the metaverse. Furthermore, our study highlights that the three distinct models utilized demonstrate varying initial defense performance against different jailbreak attack strategies, underscoring the complexity of effectively securing LLMs. © 2024 Copyright held by the owner/author(s).},
keywords = {% reductions, Attack strategies, Computer simulation languages, Defense, Digital elevation model, Guard rails, Jailbreak, Language Model, Large language model, Metaverse Security, Metaverses, Natural languages, Performance, Virtual Reality},
pubstate = {published},
tppubtype = {inproceedings}
}
Large language models (LLMs) have demonstrated exceptional capabilities in natural language processing tasks, fueling innovations in emerging areas such as the metaverse. These models enable dynamic virtual communities, enhancing user interactions and revolutionizing industries. However, their increasing deployment exposes vulnerabilities to jailbreak attacks, where adversaries can manipulate LLM-driven systems to generate harmful content. While various defense mechanisms have been proposed, their efficacy against diverse jailbreak techniques remains unclear. This paper addresses this gap by evaluating the performance of three popular defense methods (Backtranslation, Self-reminder, and Paraphrase) against different jailbreak attack strategies (GCG, BEAST, and Deepinception), while also utilizing three distinct models. Our findings reveal that while defenses are highly effective against optimization-based jailbreak attacks and reduce the attack success rate by 79% on average, they struggle in defending against attacks that alter attack motivations. Additionally, methods relying on self-reminding perform better when integrated with models featuring robust safety guardrails. For instance, Llama2-7b shows a 100% reduction in Attack Success Rate, while Vicuna-7b and Mistral-7b, lacking safety alignment, exhibit a lower average reduction of 65.8%. This study highlights the challenges in developing universal defense solutions for securing LLMs in dynamic environments like the metaverse. Furthermore, our study highlights that the three distinct models utilized demonstrate varying initial defense performance against different jailbreak attack strategies, underscoring the complexity of effectively securing LLMs. © 2024 Copyright held by the owner/author(s).