3DSynBrush a high quality 3D reconstruction framework for single Dunhuang murals
Source abstract
Abstract
The Dunhuang Mogao Grottoes are among the most significant historical and cultural sites in China, with the Dunhuang Mural serving as key representative artifacts. Due to natural erosion, human activities, and other factors, it is crucial to protect these frescoes authentically and accurately to ensure their preservation. To address this, we propose a technological framework, 3DSynBrush, for high-quality 3D reconstruction of Chinese painting data. First, single elements are extracted from the mural through the Perspective-Driven Synthesis (PDS) Module, and a sparse perspective is generated. The sparse view is then passed through the Neural Rendering Synthesizer to obtain a continuous view image, which ultimately meets the input requirements of the Light Field Fusion Meshing Module to achieve 3D reconstruction. 3DSynBrush aims to reconstruct a realistic model of the scene depicted using only a single primary view of the fresco. Moreover, we develope a high-quality Chinese Mural Elements dataset, termed CME. Compared to current 3D reconstruction algorithms, 3DSynBrush produces visually coherent results while using only 40% of the vertices and triangles, thereby minimizing computational resources.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Computer Graphics and Visualization Techniques · Image Processing and 3D Reconstruction · 3D Surveying and Cultural Heritage
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (38)
Show source-provided references
1. Lee, S. S. Repository of ingenuity: Cave 61 and artistic appropriation in tenth-century Dunhuang. Art. Bull. 94, 199–225 (2012).10.1080/00043079.2012.10786038 ↗
2. Sicheng, L. Ancient Chinese architecture as seen in the Dunhuang murals. Cult. Relics. 5, 7 (1951). (in Chinese)
3. Shengliang, Z. & Zhaofu, Q. The flying apsaras: The central pillar front niche ceiling of cave 257 in theMogao Caves, Northern Wei. Dunhuang Res. 2018, 2 (2018). (in Chinese)
4. Ru, S. The image of towers in the Dunhuang murals. Dunhuang Stud. 1996, 1–16 (1996).
5. Zhang, H. Aesthetic contemplation of the Tang Dynasty Dunhuang frescoes elements on contemporary costume design. In: Cross-Cultural Design. User Experience of Products, Services, and Intelligent Environments (ed Rau, P. L. P.) 431–440 (Lecture Notes in Computer Science, Vol. 12192, Springer, Cham, 2020).10.1007/978-3-030-49788-0_32 ↗
6. Youhui, G. Early patterns and decorations in the Dunhuang Mogao caves. Journal of Dunhuang Studies 101–107+125–126 (1980).
7. Shengliang, Z. et al. Complete Works of the Dunhuang Caves: Landscape Paintings (Commercial Press, Hong Kong, 2002).
8. Yagi, H. & Mei, L. About the Western Pure Land illustration on the south wall in the Mogao Cave 220 at Dunhuang. Dunhuang Res. 2012, 9–15 (2012).
9. Rombach, R., Blattmann, A., Lorenz, D., Esser, P., & Ommer, B. High-resolution image synthesis with latent diffusion models. 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 10674–10685 (2021).10.1109/CVPR52688.2022.01042 ↗
10. Saharia, C. et al. Photorealistic text-to-image diffusion models with deep language understanding. In Pro. 36th International Conference on Neural Information Processing Systems, NIPS ’22 (Curran Associates Inc., Red Hook, NY, USA, 2022).
11. Zhu, Y.-L., Wang, X.-Y., Wan, T.-R. & Yang, Y.-Q. The analysis and creation of Mogao Caves’ three-dimensional model. In E-Learning and Games, 191–198 (Springer International Publishing, Cham, 2017).10.1007/978-3-319-65849-0_20 ↗
12. Wang, H., Du, X., Li, J., Yeh, R. A. & Shakhnarovich, G. Score Jacobian chaining: Lifting pretrained 2d diffusion models for 3d generation. In 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 12619–12629 (2023).10.1109/CVPR52729.2023.01214 ↗
13. Tang, J. et al. Make-it-3d: High-fidelity 3d creation from a single image with diffusion prior. 2023 IEEE/CVF International Conference on Computer Vision (ICCV) 22762–22772 (2023).10.1109/ICCV51070.2023.02086 ↗
14. Zhaowen, Q. & Tianwen, Z. Key techniques on cultural relic 3d reconstruction. J. Electron. 36, 2423–2427 (2008).
15. Cheng, L., Gao, X. & Tian, X. The application of digital technology in mural preservation: A case study of Dunhuang animation. Art Style Art Cult. Int. Mag. 14, 94–111 (2024).
16. Han, P.-H. et al. A compelling virtual tour of the Dunhuang cave with an immersive head-mounted display. IEEE Comput. Graph. Appl. 40, 40–54 (2020).10.1109/MCG.2019.2936753 ↗
17. Stone, R. & Ojika, T. Virtual heritage: What next? IEEE Multimed. 7, 73–74 (2000).10.1109/93.848434 ↗
18. Liu, B., He, F., Du, S., Zhang, K. & Wang, J. Dunhuang murals contour generation network based on convolution and self-attention fusion. Appl. Intell. 53, 22073–22085 (2023).10.1007/s10489-023-04614-4 ↗
19. Chen, B., Xu, H., Yin, Z., Zhou, C. & Yang, H. Cross-scale generative adversarial learning networks for intelligent hierarchical control of proton exchange membrane fuel cells systems. Int. Commun. Heat. Mass Transf. 169, 109878 (2025).10.1016/j.icheatmasstransfer.2025.109878 ↗
20. Wang, H., Qiu, X., Xiong, Y. & Tan, X. Autogrn: An adaptive multi-channel graph recurrent joint optimization network with copula-based dependency modeling for spatio-temporal fusion in electrical power systems. Inf. Fusion 117, 102836 (2025).10.1016/j.inffus.2024.102836 ↗
21. Wang, H. et al. Rofed-llm: Robust federated learning for large language models in adversarial wireless environments. IEEE Trans. Netw. Sci. Eng. 13, 1084–1096 (2026).10.1109/TNSE.2025.3590975 ↗
22. Wang, G., Chen, Z., Loy, C. C. & Liu, Z. Sparsenerf: Distilling depth ranking for few-shot novel view synthesis. In 2023 IEEE/CVF International Conference on Computer Vision (ICCV), 9031–9042 (2023).10.1109/ICCV51070.2023.00832 ↗
23. Guo, S. et al. Depth-guided robust point cloud fusion nerf for sparse input views. IEEE Trans. Circuits Syst. Video Technol. 34, 8093–8106 (2024).10.1109/TCSVT.2024.3385360 ↗
24. Niemeyer, M. et al. Regnerf: Regularizing neural radiance fields for view synthesis from sparse inputs. In 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 5470–5480 (2022).10.1109/CVPR52688.2022.00540 ↗
25. Shi, R., Wei, X., Wang, C. & Su, H. Zerorf: Fast sparse view 360° reconstruction with zero pretraining. In 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 21114–21124 (2024).10.1109/CVPR52733.2024.01995 ↗
26. Mildenhall, B. et al. Nerf: Representing scenes as neural radiance fields for view synthesis. Commun. ACM 65, 99–106 (2021).10.1145/3503250 ↗
27. Zong, X., Li, Z. & Zhang, Q. A study on the stage image of “rebound lute behind the back” in Dunhuang, China. Int. J. Adv. Cult. Technol. 12, 16–29 (2024).
28. Dazheng, Z. Dunhuang Murals and the Colors of Chinese Painting (People’s Fine Arts Publishing House, Beijing, 2000).
29. Cheng, B., Misra, I., Schwing, A. G., Kirillov, A. & Girdhar, R. Masked-attention mask transformer for universal image segmentation. In 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 1280–1289 (2022).10.1109/CVPR52688.2022.00135 ↗
30. Liu, R. et al. Zero-1-to-3: Zero-shot one image to 3d object. 2023 IEEE/CVF International Conference on Computer Vision (ICCV) 9264–9275 (2023).10.1109/ICCV51070.2023.00853 ↗
31. Shi, R. et al. Zero123++: a Single Image to Consistent Multi-view Diffusion Base Model. arXiv e-prints arXiv:2310.15110 (2023).
32. Mildenhall, B. et al. Local light field fusion. ACM Trans. Graph. (TOG) 38, 1–14 (2019).10.1145/3306346.3322980 ↗
33. Tang, J. et al. Delicate textured mesh recovery from nerf via adaptive surface refinement. 2023 IEEE/CVF International Conference on Computer Vision (ICCV) 17693–17703 (2023).10.1109/ICCV51070.2023.01626 ↗
34. Schönberger, J. L. & Frahm, J.-M. Structure-from-motion revisited. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 4104–4113 (2016).10.1109/CVPR.2016.445 ↗
35. Lan, Y. et al. Ln3diff: Scalable latent neural fields diffusion for speedy 3d generation. In Computer Vision – ECCV 2024, 112-130 (Springer Nature Switzerland, Cham, 2025).10.1007/978-3-031-73235-5_7 ↗
36. Tang, J., Ren, J., Zhou, H., Liu, Z. & Zeng, G. Dreamgaussian: Generative Gaussian splatting for efficient 3d content creation. In Proc. International Conference on Learning Representations (ICLR, 2024).
37. Tochilkin, D. et al. TripoSR: Fast 3D object reconstruction from a single image. arXiv e-prints arXiv:2403.02151 (2024).
38. Qian, G. et al. Magic123: One image to high-quality 3d object generation using both 2d and 3d diffusion priors. In International Conference on Representation Learning, vol. 2024, 48142–48159 (2024).