International Journal of Information Technology and Computer Science (IJITCS)
IJITCS Vol. 18, No. 4, 8 Aug. 2026
Cover page and Table of Contents: PDF (size: 1051KB)
A Hybrid 3D Gaussian Splatting and Photogrammetry Framework for Industrial Virtual Reality-Based Fire Safety Training
PDF (1051KB), PP.128-141
Views: 0 Downloads: 0
Author(s)
Annanya Gali
1,*
Sneha Thombre
1
Index Terms
3D Gaussian Splatting, AI-Based Object Detection, Industrial Safety Training, Photogrammetry, Virtual Reality Systems
Abstract
In high-hazard workplaces like packaging facilities, effective fire safety is critical, but conventional practices fail to recognize actual hazards and are highly expensive to implement. This paper presents a hybrid reconstruction and artificial intelligence-driven framework that can potentially be applied to build interactive virtual reality environments. The objective of this study is to develop a scalable and cost-effective Virtual Reality based fire safety training system that balances realism and interactivity. To balance visual fidelity and interactivity, a hybrid reconstruction pipeline was developed. The complex background environment was reconstructed and rendered using 3D Gaussian Splatting, while for reconstructing key industrial objects as solid and interactive meshes, photogrammetry is used. An artificial intelligence-based system has been adopted for automatic object detection using You Only Look Once version 11 (YOLOv11) and material-based hazard classification using Bidirectional Encoder Representations from Transformers (BERT). In addition, interaction options are generated using a text generation model Fine-tuned Language Net Text-to-Text Transfer Transformer (FLAN-T5). The results indicate that the proposed framework produces high rendering capabilities with high precision, enabling efficient and scalable development of industrial safety training modules.
Cite This Paper
Annanya Gali, Sneha Thombre, "A Hybrid 3D Gaussian Splatting and Photogrammetry Framework for Industrial Virtual Reality-Based Fire Safety Training", International Journal of Information Technology and Computer Science(IJITCS), Vol.18, No.4, pp.128-141, 2026. DOI:10.5815/ijitcs.2026.04.08
Reference
[1]Naranjo J. E. et al., “A scoping review on virtual reality-based industrial training,” Applied Sciences, vol. 10, no. 22, article 8224, 2020, doi:10.3390/app10228224.
[2]Vercelli G. et al., “From risk to readiness: VR-based safety training for industrial hazards,” arXiv preprint arXiv:2412.13725, 2024, doi:10.48550/arXiv.2412.13725.
[3]Chen S.-Y. and Chien W.-C., “Immersive virtual reality serious games with DL-assisted learning in high-rise fire evacu- ation on fire safety training and research,” Frontiers in Psychology, vol. 13, article 786314, May 2022, doi:10.3389/fp- syg.2022.786314.
[4]Shahu A., Kinzer K., and Michahelles F., “Enhancing professional training with single-user virtual reality: Unveiling challenges, opportunities, and design guidelines,” in Proceedings of the 22nd International Conference on Mobile and Ubiquitous Multimedia (MUM ’23), 2023, pp. 244–256, doi:10.1145/3626705.3627791.
[5]Fu Y. and Li Q., “A virtual reality-based serious game for fire safety behavioral skills training,” International Journal of Human–Computer Interaction, vol. 40, no. 19, pp. 5980–5996, 2024, doi: 10.1080/10447318.2023.2247585.
[6]Faiz T. et al., “A scoping review on hazard recognition and prevention using augmented and virtual reality,” Computers, vol. 13, no. 12, article 307, 2024, doi:10.3390/computers13120307.
[7]Kerbl B. et al., “3D Gaussian splatting for real-time radiance field rendering,” ACM Transactions on Graphics, vol. 42, no. 4, July 2023, doi:10.48550/arXiv.2308.04079.
[8]Li C. et al., “RT-NeRF: Real-time on-device neural radiance fields towards immersive AR/VR rendering,” in Proceedings of the 2022 IEEE/ACM International Conference on Computer-Aided Design (ICCAD ’22), San Diego, CA, USA, 2022, 9 pages, doi: 10.1145/3508352.3549380.
[9]Neff T. et al., “DONeRF: Towards real-time rendering of compact neural radiance fields using depth oracle networks,” Computer Graphics Forum, vol. 40, no. 4, pp. 45–59, 2021, doi:10.1111/cgf.14340.
[10]Samudrala S., Kondguli S., and Gratz P., “Benchmarking 3D Gaussian splatting rendering,” in Proceedings of the 2025 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS), Ghent, Belgium, 2025,
pp. 227–238, doi:10.1109/ISPASS64960.2025.00029.
[11]Niemeyer M. et al., “RadSplat: Radiance field-informed Gaussian splatting for robust real-time rendering with 900+ FPS,” arXiv preprint arXiv:2403.13806, 2024, doi:10.48550/arXiv.2403.13806.
[12]Gue´don A. and Lepetit V., “SuGaR: Surface-aligned Gaussian splatting for efficient 3D mesh reconstruction and high- quality mesh rendering,” arXiv preprint arXiv:2311.12775, 2023, doi: 10.48550/arXiv.2311.12775.
[13]Y. Jiang et al., “VR-GS: A physical dynamics-aware interactive Gaussian splatting system in VR,” Proc. ACM Comput. Graph. Interact. Tech., vol. 7, no. 1, art. no. 12, May 2024, doi:10.1145/3649930.
[14]Xie T. et al., “PhysGaussian: Physics-integrated 3D Gaussians for generative dynamics,” arXiv preprint arXiv:2311.12198, 2023, doi:10.48550/arXiv.2311.12198.
[15]Ye T., Huang H., Liu Y., and Yang J., “Global structure-from-motion enhanced neural radiance fields 3D reconstruction,” International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, vol. XLVIII-4/W10- 2024, pp. 199–204, 2024, doi:10.5194/isprs-archives-XLVIII-4-W10-2024-199-2024.
[16]Ahmad J. et al., “From fields to splats: A cross-domain survey of real-time neural scene representations,” arXiv preprint arXiv:2509.23555, 2025, doi: 10.48550/arXiv.2509.23555.
[17]Liu T. et al., “MVSGaussian: Fast generalizable Gaussian splatting reconstruction from multi-view stereo,” arXiv preprint arXiv:2405.12218, 2024, doi: 10.48550/arXiv.2405.12218.
[18]Montas-Laracuente N. et al., “Automatic 3D reconstruction: Mesh extraction based on Gaussian splatting from Ro- manesque–Mude´jar churches,” Applied Sciences, vol. 15, no. 15, article 8379, 2025, doi: 10.3390/app15158379.
[19]Geng X. et al., “YOLOFM: An improved fire and smoke object detection algorithm based on YOLOv5n,” Scientific Reports, vol. 14, article 4543, 2024, doi: 10.1038/s41598-024-55232-0.
[20]Alkhammash E. H., “A comparative analysis of YOLOv9, YOLOv10, YOLOv11 for smoke and fire detection,” Fire, vol. 8, no. 1, article 26, 2025, doi: 10.3390/fire8010026.
[21]Chetoui M. and Akhloufi M. A., “Fire and smoke detection using fine-tuned YOLOv8 and YOLOv7 deep models,” Fire, vol. 7, no. 4, article 135, 2024, doi: 10.3390/fire7040135.
[22]J. S. Almeida, C. Huang, F. G. Nogueira, S. Bhatia, and V. H. C. de Albuquerque, “EdgeFireSmoke: A novel lightweight CNN model for real-time video fire–smoke detection,” IEEE Trans. Ind. Informatics, vol. 18, no. 11, pp. 7889–7898, Nov. 2022, doi: 10.1109/TII.2021.3138752.
[23]Thai H.-T., Tran-Van N.-Y., Le-Minh K.-H., and Le K.-H., “An edge-based fire detection system for real-time IoT applications,” in Proceedings of the 2023 IEEE International Conference on Communication, Networks and Satellite (COMNETSAT), Malang, Indonesia, 2023, pp. 646–651, doi: 10.1109/COMNETSAT59769.2023.10420588.
[24]Nadeem M. et al., “Visual intelligence in smart cities: A lightweight deep learning model for fire detection in an IoT environment,” Smart Cities, vol. 6, no. 5, pp. 2245–2259, 2023, doi: 10.3390/smartcities6050103
[25]Mao H. et al., “LIVE-GS: LLM powers interactive VR by enhancing Gaussian splatting,” arXiv preprint arXiv:2412.09176, 2024, doi: 10.48550/arXiv.2412.09176.
[26]N. Khan, K. Muhammad, T. Hussain, M. Nasir, M. Munsif, A. S. Imran, and M. Sajjad, “An adaptive game-based learning strategy for children road safety education and practice in virtual space,” Sensors, vol. 21, no. 11, article 3661, 2021, doi: 10.3390/s21113661
[27]H. W. Chung et al., “Scaling instruction-finetuned language models,” arXiv preprint arXiv:2210.11416, 2022.
[28]Devlin J., Chang M.-W., Lee K., and Toutanova K., “BERT: Pre-training of deep bidirectional transformers for language understanding,” in Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT), 2019, pp. 4171–4186, doi: 10.18653/v1/N19-1423.
[29]Barron J. T., Mildenhall B., Tancik M., Hedman P., Martin-Brualla R., and Srinivasan P. P., “Mip-NeRF: A multiscale representation for anti-aliasing neural radiance fields,” arXiv preprint arXiv:2103.13415, 2021, doi: 10.48550/arXiv.2103.13415.
[30]Mu¨ller T., Evans A., Schied C., and Keller A., “Instant neural graphics primitives with a multiresolution hash encoding,” ACM Transactions on Graphics, vol. 41, no. 4, article 102, 15 pages, July 2022, doi: 10.1145/3528223.3530127.