RAME Logo

RAME PUBLISHERS

A better space for Quality Research
International Journal of Computational and Electronics Aspects in Engineering (IJCEAE)
Volume 7 · Issue 5 · November 2026 · pp. 562–575
Research Article · Peer Reviewed
Received: August 5, 2026 · Accepted: October 9, 2026 · Published: November 1, 2026
DOI: 10.26706/ijceae.7.5.20261003
Open Access · CC BY 4.0

Design and Implementation of a Preprocessing-Aware Multi-Input AI Framework for 3D Object Generation

Hussam Ahmed Rammo1,*, Mays Q. Sedeeq2, Razan Abdulhammed3, Shamil Khudhur Ramadhan4, Fahad Ahmed Shaban5
1Technical Engineering College for Computer and Artificial Intelligence, Northern Technical University, Nineveh, Mosul, Iraq.
2Department of Cybersecurity Engineering, College of Computer Engineering and Artificial Intelligence, Northern Technical University (NTU), Mosul, Iraq.
3Artificial Research Center, Northern Technical University, Nineveh, Iraq.
4Department of Clinical Chemistry and Artificial Intelligence, High Institute of Health Technologies and Artificial Intelligence, Northern Technical University, Mosul, Iraq.
5Department of Computer Networks and Internet, College of Information Technology, Ninevah University, Mosul, Iraq.

*Correspondence: hussam.rammo@ntu.edu.iq
Download PDF Cite this article Certificate Volume 7 Issue 5

Abstract

Although the technical challenges to create 3D content have been lowered by the application of artificial intelligence, practical challenges remain with heterogeneous input types, artefacts from background, computational demands, and integration of the system. This paper proposes a preprocessing-aware multi-input method for creating textured 3D objects from text, image and sketch input in a single application with a unified client server system. The text-to-3D workflow is implemented using FLUX.1-schnell to create an intermediate image, and Hunyuan3D-2.1 is used for the 3D shape and texture synthesis of images generated by text, photographs and sketches.

The front end is designed using Flutter, and the backend is developed using Python FastAPI, which handles the processing of pre-processing, model inference, job monitoring and generation of GLB output, while heavy computation is performed remotely on an NVIDIA A100 80 GB GPU. The structure contains background removal, alpha thresholding, object cropping and centred square padding to minimize geometry from background. As a result of functional testing, the successful generation and in-application preview were confirmed for all three workflows. The end-to-end generation times were 80.72 s (sketch-to-3D), 103.89 s (image-to-3D), and ~105 s (text-to-3D) respectively. In the evaluated case, the ground plane artefact was observed when preprocessing was performed using background removal only. For the case in question, a qualitative pre-processing ablation indicated that the observed artefact on the ground plane can be removed with an object-only cleaning. The results show that the current generative models, pre-processing, and inference on cloud can be integrated within a unified 3D generation pipeline. The current assessment is still at prototype level due to lack of repeated-trial statistics, ground truth geometric standards and wider validation.

Keywords

3D object generation multi-input framework text-to-3D image-to-3D sketch-to-3D input preprocessing

References

  1. A. Nichol, H. Jun, P. Dhariwal, P. Mishkin, and M. Chen, “Point-e: A system for generating 3d point clouds from complex prompts,” arXiv preprint arXiv:2212.08751, 2022.
  2. T. Hunyuan3D et al., “Hunyuan3d 2.1: From images to high-fidelity 3d assets with production-ready pbr material,” arXiv preprint arXiv:2506.15442, 2025.
  3. H. Qin et al., “AViTS: Adaptive Spatiotemporal Token Selection for Efficient Dynamic-Resolution Generation,” in European Conference on Computer Vision (ECCV), Springer, 2026, pp. 257–273.
  4. B. Poole, A. Jain, J. T. Barron, and B. Mildenhall, “Dreamfusion: Text-to-3d using 2d diffusion,” arXiv preprint arXiv:2209.14988, 2022.
  5. C.-H. Lin et al., “Magic3d: High-resolution text-to-3d content creation,” in Proc. 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, 2023, pp. 300–309.
  6. Z. Wang et al., “Prolificdreamer: High-fidelity and diverse text-to-3d generation with variational score distillation,” Adv. Neural Inf. Process. Syst., vol. 36, pp. 8406–8441, 2023.
  7. Y. Shi, P. Wang, J. Ye, L. Mai, K. Li, and X. Yang, “Mvdream: Multi-view diffusion for 3d generation,” in Proc. International Conference on Learning Representations (ICLR), 2024, pp. 39838–39859.
  8. H. Jun and A. Nichol, “Shap-e: Generating conditional 3d implicit functions,” arXiv preprint arXiv:2305.02463, 2023.
  9. T. Yi et al., “Gaussiandreamer: Fast generation from text to 3d gaussians by bridging 2d and 3d diffusion models,” in Proc. 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, 2024, pp. 6796–6807.
  10. H. Lee, M. Savva, and A. X. Chang, “Text‐to‐3D Shape Generation,” Computer Graphics Forum, Wiley Online Library, vol. 43, no. 2, p. e15061, 2024.
  11. J. Liu et al., “A comprehensive survey on 3d content generation,” arXiv preprint arXiv:2402.01166, 2024.
  12. R. Liu, R. Wu, B. Van Hoorick, P. Tokmakov, S. Zakharov, and C. Vondrick, “Zero-1-to-3: Zero-shot one image to 3d object,” in Proc. 2023 IEEE/CVF International Conference on Computer Vision (ICCV), IEEE, 2023, pp. 9264–9275.
  13. Y. Hong et al., “Lrm: Large reconstruction model for single image to 3d,” in Proc. International Conference on Learning Representations (ICLR), 2024, pp. 50678–50702.
  14. J. Xu, W. Cheng, Y. Gao, X. Wang, S. Gao, and Y. Shan, “Instantmesh: Efficient 3d mesh generation from a single image with sparse-view large reconstruction models,” arXiv preprint arXiv:2404.07191, 2024.
  15. X. Long et al., “Wonder3d: Single image to 3d using cross-domain diffusion,” in Proc. 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, 2024, pp. 9970–9980.
  16. D. Tochilkin et al., “Triposr: Fast 3d object reconstruction from a single image,” arXiv preprint arXiv:2403.02151, 2024.
  17. L. Chen et al., “3D-ISRNet: Generating 3D point clouds through image similarity retrieval in a complex background from a single image,” Image Vis. Comput., vol. 150, p. 105203, 2024.
  18. Y. Tong, H. Chen, N. Yang, M. I. Menhas, and B. Ahmad, “3D-CDRNet: Retrieval-based dense point cloud reconstruction from a single image under complex background,” Displays, vol. 78, p. 102438, 2023.
  19. Y. Yang and B. Yang, “Benchmarking and analysis of unsupervised object segmentation from real-world single images,” Int. J. Comput. Vis., vol. 132, no. 6, pp. 2077–2113, 2024.
  20. W. Cai et al., “Instruct Pix-to-3D: Instructional 3D object generation from a single image,” Neurocomputing, vol. 600, p. 128156, 2024.
  21. S. K. Ramadhan, N. R. Najem, A. Sulaiman, H. A. Rammo, and F. A. Shaban, “Beyond Chronological Age: Explainable Multi-Task Swin Transformers for Biological Age and Gender Estimation from Dorsal Hand Images,” Int. J. Comput. Electron. Aspects Eng., vol. 7, no. 2, pp. 227–243, 2026.
  22. J. Xiang et al., “Structured 3d latents for scalable and versatile 3d generation,” in Proc. 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, 2025, pp. 21469–21480.
  23. Z. Zhao et al., “Hunyuan3d 2.0: Scaling diffusion models for high resolution textured 3d assets generation,” arXiv preprint arXiv:2501.12202, 2025.
  24. T. Chen et al., “Deep3dsketch+: Rapid 3d modeling from single free-hand sketches,” in Proc. International Conference on Multimedia Modeling, Springer, 2023, pp. 16–28.
  25. P. Esser et al., “Scaling rectified flow transformers for high-resolution image synthesis,” in Proc. Forty-first International Conference on Machine Learning (ICML), 2024.
  26. C. Wang, H.-Y. Peng, Y.-T. Liu, J. Gu, and S.-M. Hu, “Diffusion models for 3D generation: A survey,” Comput. Vis. Media, vol. 11, no. 1, pp. 1–28, 2025.
  27. Flutter Documentation, “Flutter: Build apps for any screen,” Google LLC, 2026. [Online]. Available: https://docs.flutter.dev.
  28. P. Pataranutaporn, A. Doudkin, and P. Maes, “OceanChat: The effect of virtual conversational AI agents on sustainable attitude and behavior change,” in Proc. 2026 CHI Conference on Human Factors in Computing Systems, 2026, pp. 1–25.
  29. S. Tiangolo, “FastAPI documentation,” 2025. [Online]. Available: https://fastapi.tiangolo.com.
  30. L. Körösi et al., “From Detection to Action: Implementing Deep Learning Inference in PLC Systems via Docker Services,” IFAC-PapersOnLine, vol. 58, no. 9, pp. 19–24, 2024.