Intelligent generation of Peking opera facial masks with deep learning frameworks

Heritage Science

Research

Yan Wang¹  Rui Xiong²  Yinghua Shen²  Cong Jin²  Ming Yan³ 

[1] School of Data Science and Intelligent Media, Communication University of China, 100024, Beijing, China;School of Information and Communication Engineering, Communication University of China, 100024, Beijing, China;School of Information and Communication Engineering, Communication University of China, 100024, Beijing, China;Key Laboratory of Acoustic Visual Technology and Intelligent Control System, Ministry of Culture and Tourism, 100024, Beijing, China;

关键词: Peking opera facial masks;    Image generation;    Generative adversarial network (GAN);    Image translation;   

DOI  :  10.1186/s40494-023-00865-z

 received in 2022-12-07, accepted in 2023-01-13,  发布年份 2023

来源: Springer

【 摘 要 】

The production of traditional Peking opera facial masks often relies on hand painting by experienced painters, which restricts the inheritance and development of this intangible cultural heritage. Current research mainly focuses on the digital reconstruction and storage of existing Peking opera facial masks, while high-quality facial mask generation technology is still in an infancy stage. In this paper, different deep learning frameworks are improved for learning features of Peking opera facial masks and generating new masks, which can effectively promote the creative application of Peking opera facial masks. First, using different data enhancement methods, an improved Style Generative Adversarial Network-2 (StyleGAN2) can learn implicit and explicit features of Peking opera facial masks and automatically generate new facial masks. In addition, an image translation framework for joint cross-domain communication under weak supervision is used to translate face sketches and color reference maps to an intermediate feature domain, and then synthesize new facial masks through an image generation network. The experimental results show that the generated Peking opera facial masks have good local randomness and excellent visual quality.

【 授权许可】

CC BY   
© The Author(s) 2023

【 图 表 】

Fig. 1

Fig. 4

Fig. 1

41116_2022_35_Article_IEq71.gif

12888_2022_4506_Article_IEq2.gif

Fig. 1

12888_2022_4506_Article_IEq1.gif

Fig. 4

Fig. 1 

41116_2022_35_Article_IEq68.gif

Fig. 5

41116_2022_35_Article_IEq66.gif

40854_2022_443_Article_IEq20.gif

Fig. 4

41116_2022_35_Article_IEq65.gif

41116_2022_35_Article_IEq64.gif

41116_2022_35_Article_IEq63.gif

41116_2022_35_Article_IEq62.gif

41116_2022_35_Article_IEq61.gif

41116_2022_35_Article_IEq60.gif

Fig. 4

41116_2022_35_Article_IEq59.gif

Fig. 2

41116_2022_35_Article_IEq58.gif

Fig. 2

41116_2022_35_Article_IEq57.gif

Fig. 1

Fig. 1

41116_2022_35_Article_IEq56.gif

41116_2022_35_Article_IEq55.gif

Fig. 3

Fig. 3

Fig. 4

41116_2022_35_Article_IEq53.gif

41116_2022_35_Article_IEq52.gif

41116_2022_35_Article_IEq51.gif

40249_2022_1049_Article_IEq2.gif

41116_2022_35_Article_IEq50.gif

Fig. 2

41116_2022_35_Article_IEq49.gif

Fig. 4

40249_2022_1049_Article_IEq1117.gif

Fig. 1

Fig. 12

40249_2022_1049_Article_IEq39.gif

40249_2022_1049_Article_IEq33.gif

40249_2022_1049_Article_IEq32.gif

Fig. 7

Fig. 1

Fig. 1

Fig. 1

41116_2022_35_Article_IEq40.gif

41116_2022_35_Article_IEq39.gif

41116_2022_35_Article_IEq38.gif

41116_2022_35_Article_IEq37.gif

40249_2022_1049_Article_IEq26.gif

Fig. 5

Fig. 3

40249_2022_1049_Article_IEq21.gif

40249_2022_1049_Article_IEq57.gif

Fig. 5

Fig. 1

Fig. 6

41116_2022_35_Article_IEq15.gif

41116_2022_35_Article_IEq7.gif

Fig. 2

41116_2022_35_Article_IEq5.gif

41116_2022_35_Article_IEq4.gif

Fig. 1

Fig. 3

Fig. 3

42004_2022_800_Article_IEq108.gif

Fig. 1

Fig. 5

Fig. 3

Fig. 2

Fig. 3

42004_2022_803_Article_IEq37.gif

Fig. 5

Fig. 62

13690_2022_1010_Article_IEq1.gif

40798_2022_490_Article_IEq48.gif

40798_2022_490_Article_IEq44.gif

Fig. 4