Traditional Chinese textile patterns are often used as decorative elements on garments, fabrics, and accessories, adding unique cultural charm and aesthetic value to textile products. Recently, the big advances in data-driven artificial intelligence (AI) technology offer new ways to study textile patterns, but collecting enough data for training is challenging. One-shot image synthesis technology can help in constructing datasets under data-scarce conditions. Nevertheless, existing one-shot image generation methods are inadequate in fulfilling the structural layout and efficiency requirements for the production of traditional Chinese textile patterns. This paper proposes a one-shot Chinese textile pattern image generation (OS-CTPIG) method, which facilitates the generation of Chinese textile patterns with central radial structures, faithful colors, and exquisite details. A multiscale framework is employed for the generation of details in a coarse-to-fine manner. The simultaneous control over color and structure, along with the respective optimization at the color space and spatial feature space, complemented by a color leakage constraint module, is designed to preserve the traditional aesthetic standards in structures. Moreover, the integration of an energy function-enhanced sliced Wasserstein distance optimization objective and a dynamic task-weighting mechanism is proposed to improve computational efficiency. The experimental results demonstrate that the OS-CTPIG method is both effective and efficient in generating traditional Chinese textile pattern images under one-shot conditions. Subsequently, a dataset comprising over 3000 images of traditional Chinese textile pattern images has been constructed successfully.
LiMWangYXuY-Q.Computing for Chinese cultural heritage. Vis Informat2022;
6: 1–13.
2.
FanTWangH.Research of Chinese intangible cultural heritage knowledge graph construction and attribute value extraction with graph attention network. Informat Process Managem2022;
59: 102753.
3.
FiorucciMKhoroshiltsevaMPontilM, et al.
Machine learning for cultural heritage: A survey. Pattern Recog Lett2020;
133: 102–108.
4.
WuQZhuBYongB, et al.
ClothGAN: Generation of fashionable Dunhuang clothes using generative adversarial networks. Connect Sci2021;
33: 341–358.
5.
ChenLSuZHeX, et al.
The application of robotics and artificial intelligence in embroidery: Challenges and benefits. Assembly Automat2022;
42: 851–868.
6.
KumarVHernándezNJensenM, et al.
Deep learning based system for garment visual degradation prediction for longevity. Comput Ind2023;
144: 103779.
7.
ZengZSunSLiT, et al.
Exploring the topic evolution of Dunhuang murals through image classification. J Informat Sci2024;
50: 35–52.
8.
ZhouZLiuXShangJ, et al.
Inpainting digital dunhuang murals with structure-guided deep network. ACM J Comput Cult Heritage2022;
15: 1–25.
9.
WangJLiJLiuW, et al.
Dunhuang mural line drawing based on multi-scale feature fusion and sharp edge learning. Neur Process Lett2023;
55: 10201–10214.
10.
SunKZhangPZhangJ, et al.
PFNet: Attribute-aware personalized fashion editing with explainable fashion compatibility analysis. Informat Process Managem2024;
61: 103540.
11.
GuoY.Design of artistic graphic symbols based on intelligent guidance marking system. Neur Comput Applicat2023;
35: 4255–4266.
12.
HeynH-MKnaussEMalleswaranI, et al.
An empirical investigation of challenges of specifying training data and runtime monitors for critical software with machine learning and their relation to architectural decisions. Requirements Eng2024;
29: 97–117.
13.
AbdollahzadehMMalekzadehTTeoCTH, et al. A survey on generative modeling with limited data, few shots, and zero shot, http://arxiv.org/abs/2307.14397 (2023, accessed 7 June 2024).
14.
LvCZhangDGengS, et al.
Color transfer for images: A survey. ACM Trans Multimedia Comput Commun Appl2023; 3635152.
15.
WangYGonzalez-GarciaABergaD, et al. MineGAN: Effective knowledge transfer from GANs to target domains with few images. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 9332–9341.
16.
PhaphuangwittayakulAGuoYYingF.Fast adaptive meta-learning for few-shot image generation. IEEE Trans Multimedia2021;
24: 2205–2217.
17.
YangMWangZ. Image synthesis under limited data: A survey and taxonomy, https://arxiv.org/abs/2307.16879 [cs] (2023, accessed 27 March 2024).
18.
SushkoVGallJKhorevaA. One-shot GAN: Learning to generate samples from single images and videos. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 2596–2600.
19.
KirillovAMintunERaviN, et al. Segment anything. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 4015–4026.
20.
BonneelNRabinJPeyréG, et al.
Sliced and radon Wasserstein barycenters of measures. J Math Imaging Vision2015;
51: 22–45.
21.
KarrasTAittalaMHellstenJ, et al.
Training generative adversarial networks with limited data. Adv Neur Informat Process Syst2020;
33: 12104–12114.
22.
JiangLDaiBWuW, et al.
Deceive d: Adaptive pseudo augmentation for gan training with limited data. Advances in Neural Information Processing Systems2021;
34: 21655–21667.
23.
LiZUsmanMTaoR, et al.
A systematic survey of regularization and normalization in GANs. ACM Comput Surv2023;
55: 1–37.
24.
LiuBZhuYSongK, et al. Towards faster and stabilized GAN training for high-fidelity few-shot image synthesis. In: International Conference on Learning Representations, 2020.
25.
ZhaoMCongYCarinL. On leveraging pretrained GANs for generation with limited data. In: International conference on Machine Learning. PMLR, 2020, pp. 11340–11351.
26.
MoonTChoiMLeeG, et al. Fine-tuning diffusion models with limited data. In: NeurIPS 2022 Workshop on Score-Based Methods, 2022.
27.
HospedalesTAntoniouAMicaelliP, et al.
Meta-learning in neural networks: A survey. IEEE Trans Pattern Anal Machine Intell2021;
44: 5149–5169.
28.
HongYNiuLZhangJ, et al. Few-shot image generation using discrete content representation. In: Proceedings of the 30th ACM International Conference on Multimedia, 2022, pp. 2796–2804.
29.
GuZLiWHuoJ, et al. LofGAN: Fusing local representations for few-shot image generation. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 8463–8471.
30.
ShahamTRDekelTMichaeliT. SinGAN: Learning a generative model from a single natural image. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 4570–4580.
31.
ZhangZHanCGuoT. ExsinGAN: Learning an explainable generative model from a single image. arXiv preprint arXiv:210507350.
32.
WangWBaoJZhouW, et al. Sindiffusion: Learning a diffusion model from a single natural image. arXiv preprint arXiv:221112445.
33.
NikankinYHaimNIraniM. Sinfusion: Training diffusion models on a single image or video. arXiv preprint arXiv:221111743.
34.
GranotNFeinsteinBShocherA, et al. Drop the GAN: In defense of patches nearest neighbors as single image generative models. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 13460–13469.
35.
ElnekaveAWeissY. Generating natural images with direct patch distributions matching. In: European Conference on Computer Vision. Berlin: Springer, 2022, pp. 544–560.
36.
NietertSGoldfeldZSadhuR, et al.
Statistical, robustness, and computational guarantees for sliced Wasserstein distances. Adv Neur Informat Process Syst2022;
35: 28179–28193.
37.
LeDNguyenHNguyenK, et al. Fast approximation of the generalized sliced-Wasserstein distance. In: ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Piscataway, NJ: IEEE, 2024, pp. 6920–6924.
38.
KolouriSNadjahiKSimsekliU, et al.
Generalized sliced Wasserstein distances. Adv Neur Informat Process Syst2019. https://proceedings.neurips.cc/paper_files/paper/2019/hash/f0935e4cd5920aa6c7c996a5ee53a70f-Abstract.html
39.
DeshpandeIHuY-TSunR, et al. Max-sliced Wasserstein distance and its use for GANs. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, pp. 10648–10656.
40.
XiJNiles-WeedJ.Distributional convergence of the sliced Wasserstein process. Adv Neur Informat Process Syst2022;
35: 13961–13973.
41.
BonetCBergPCourtyN, et al. Spherical sliced-Wasserstein. arXiv preprint arXiv:220608780.
42.
CrawshawM. Multi-task learning with deep neural networks: A survey. arXiv preprint arXiv:200909796.
43.
ChenZBadrinarayananVLeeC-Y, et al. Gradnorm: Gradient normalization for adaptive loss balancing in deep multitask networks. In: International Conference on Machine Learning. PMLR, 2018, pp. 794–803.
44.
YuTKumarSGuptaA, et al.
Gradient surgery for multi-task learning. Adv Neur Informat Process Syst2020;
33: 5824–5836.
45.
LiuBFengYStoneP, et al.
FAMO: Fast adaptive multitask optimization. Adv Neur Informat Process Syst2023;
36: 57226–57243.
46.
ZhaoXLiHShenX, et al. A modulation module for multi-task learning with applications in image retrieval. In: Proceedings of the European Conference on Computer Vision (ECCV), 2018, pp. 401–416.
47.
ChenZNgiamJHuangY, et al.
Just pick a sign: Optimizing deep multitask models with gradient sign dropout. Adv Neur Informat Process Syst2020;
33: 2039–2050.
48.
SanhVWolfTRuderS. A hierarchical multi-task approach for learning embeddings from semantic tasks. In: Proceedings of the AAAI Conference on Artificial Intelligence, 2019, pp. 6949–6956.
FuSHamiltonMBrandtL, et al. Featup: A model-agnostic framework for features at any resolution. arXiv preprint arXiv:240310516.
51.
OquabMDarcetTMoutakanniT, et al. Dinov2: Learning robust visual features without supervision. arXiv preprint arXiv:230407193.
52.
HanKWangYChenH, et al.
A survey on vision transformer. IEEE Trans Pattern Anal Machine Intell2022;
45: 87–110.
53.
MittalAMoorthyAKBovikAC.No-reference image quality assessment in the spatial domain. IEEE Trans Image Process2012;
21: 4695–4708.
54.
WangJChanKCLoyCC. Exploring clip for assessing the look and feel of images. In: Proceedings of the AAAI Conference on Artificial Intelligence, 2023, pp. 2555–2563.
55.
MaCYangC-YYangX, et al.
Learning a no-reference quality metric for single-image super-resolution. Comput Vis Image Understand2017;
158: 1–16.
56.
TalebiHMilanfarP.NIMA: Neural image assessment. IEEE Trans Image Process2017;
27(8): 3998–4011.
57.
ZhangZLiuYHanC, et al. PetsGAN: Rethinking priors for single image generation. In: Proceedings of the AAAI Conference on Artificial Intelligence, 2022, pp. 3408–3416.
58.
ZhangYTangFDongW, et al. Domain enhanced arbitrary image style transfer via contrastive learning. In: Special Interest Group on Computer Graphics and Interactive Techniques Conference Proceedings, pp. 1–8.
59.
ChenHZhaoLLiJ, et al. TSSAT: Two-stage statistics-aware transformation for artistic style transfer. In: Proceedings of the 31st ACM International Conference on Multimedia, 2023.
60.
BhattacharyyaA.On a measure of divergence between two statistical populations defined by their probability distribution. Bull Calcutta Math Soc1943;
35: 99–110.
61.
ZhaoF.Ornamental Patterns from Ancient Chinese Textiles: Han to Tang Dynasties.
Zhejiang University Press, 2018, pp. 10–108.
