Contour-enhanced CycleGAN framework for style transfer from scenery photos to Chinese landscape paintings

Peng, Xianlin; Peng, Shenglin; Hu, Qiyao; Peng, Jinye; Wang, Jiaxin; Liu, Xinyu; Fan, Jianping

doi:10.1007/s00521-022-07432-w

Cited by 16 publications

(7 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Subsequently, GAN-based coloring methods were applied to painting tasks. For instance, Peng and others employed the CE-CycleGAN framework to process edge contours and transform landscape photos into Chinese landscape painting styles [11]. Sun and others developed a system capable of fitting edges based on semantic label maps [12], generating exquisite paintings of a specific type.…”

Section: Image Coloringmentioning

confidence: 99%

Enhancing Urban Landscape Design: A GAN-Based Approach for Rapid Color Rendering of Park Sketches

Chen,

Zhao,

Yao

et al. 2024

Land

View full text Add to dashboard Cite

In urban ecological development, the effective planning and design of living spaces are crucial. Traditional color plan rendering methods, mainly using generative adversarial networks (GANs), rely heavily on edge extraction. This often leads to the loss of important details from hand-drawn drafts, significantly affecting the portrayal of the designer’s key concepts. This issue is especially critical in complex park planning. To address this, our study introduces a system based on conditional GANs. This system rapidly converts black-and-white park sketches into comprehensive color designs. We also employ a data augmentation strategy to enhance the quality of the output. The research reveals: (1) Our model efficiently produces designs suitable for industrial applications. (2) The GAN-based data augmentation improves the data volume, leading to enhanced rendering effects. (3) Our unique approach of direct rendering from sketches offers a novel method in urban planning and design. This study aims to enhance the rendering aspect of an intelligent workflow for landscape design. More efficient rendering techniques will reduce the iteration time of early design solutions and promote the iterative speed of designers’ thinking, thus improving the speed and efficiency of the whole design process.

show abstract

Section: Image Coloringmentioning

confidence: 99%

Enhancing Urban Landscape Design: A GAN-Based Approach for Rapid Color Rendering of Park Sketches

Chen,

Zhao,

Yao

et al. 2024

Land

View full text Add to dashboard Cite

show abstract

“…The outcomes demonstrated that the framework can enhance the landscape painting effect of landscape photographs with a comprehensive similarity score as high as 0.92. In the comparative analysis, the method outperformed several existing reference models in terms of visual quality [7]. Zhou et al improved the traditional codebook modeling algorithm and proposed an improved codebook modeling algorithm.…”

Section: Related Workmentioning

confidence: 99%

“…the true value coordinates and predicted value coordinates, respectively. Equation ( 6) is viewed as a whole for regression, and its calculation equation is obtained as shown in equation (7)…”

Section: ( )mentioning

confidence: 99%

Real-time Target Detection System in Scenic Landscape Based on Improved YOLOv4 Algorithm

Pan,

Zhao,

Sun

2024

IJCAI

View full text Add to dashboard Cite

With the rapid development of computer vision technology, the use of real-time target detection systems in scenic landscape management and services is increasingly widespread. To enhance the precision and efficiency of real-time target detection in scenic landscapes, this research integrates the fourth version of the You Only Look Once (YOLO) algorithm to construct an optimized real-time target detection system is introduced for scenic landscapes. First, adaptive spatial feature fusion to enhance the fourth version of the You Only Look Once algorithm. Then, the optimized algorithm was combined with OpenCV library, Python OS library, and other hardware and software to design a real-time image recognition system for scenic landscapes. The study results indicated that the proposed optimized algorithm had better recognition performance, and its precision value, recall rate, and F1 value were as high as 0.96, 0.97, and 0.98, respectively. The recognition system, which was developed using an optimization algorithm, demonstrated excellent practical application effect. It displayed stable system operation under four natural landscapes: sunrise, sea of clouds, maple forest, and stone monument, with a stability performance of 0.92, 0.93, 0.92, and 0.94, respectively. Moreover, the system operated remarkably fast, with low operational times of 2.3 s, 0.8 s, 2.9 s, and 1.2 s under these landscapes. In conclusion, the research institute's target detection algorithm has demonstrated excellent performance. Utilizing this algorithm in the detection system can offer technical aid for managing and intelligently detecting scenic landscape images.Povzetek: Raziskava predstavlja izboljšan sistem za zaznavanje tarč v realnem času v slikovitih krajinskih območjih, temelječ na algoritmu YOLOv4, s prilagodljivo prostorsko združitvijo značilnosti in uporabo knjižnic OpenCV in Python OS.

show abstract

“…A U-Net skip structure is added to the network structure of the generator [10][11] to preserve pixel level image information at different resolutions. As shown in Figure 5, the U-Net network structure diagrams.…”

Section: Improvements To the Cyclegan Network Structurementioning

confidence: 99%

“…In formula (11), the latter half is the canonical term of the spectral norm of the weight matrix, and the discriminator satisfies the Lipschitz constraint by punishing the sum of the spectral norm of each layer.…”

Section: ) Introduction Of Wasserstein Distance Instead Of the Origin...mentioning

confidence: 99%

Research on Data Enhancement Method for Complex Pavement

2023

IJNDES

View full text Add to dashboard Cite

In the era of intelligent vehicle development, environmental awareness technology as the basis of vehicle decision-making and route planning, affected by the weather, accurate road image recognition is particularly important. Aiming at this problem, this paper constructs a large scale pavement image dataset and proposes an improved image generation method based on CycleGAN data enhancement. The core innovation of the method is the improvement of the model structure, introduce Wassertein distance, and implement global Lipschitz constraint with spectral constraint. The experimental results demonstrate that the accuracy of the data set is up to 91.85%, which is 6.41% higher than the original CycleGAN, and the convergence speed is faster.

show abstract

Contour-enhanced CycleGAN framework for style transfer from scenery photos to Chinese landscape paintings

Cited by 16 publications

References 23 publications

Enhancing Urban Landscape Design: A GAN-Based Approach for Rapid Color Rendering of Park Sketches

Enhancing Urban Landscape Design: A GAN-Based Approach for Rapid Color Rendering of Park Sketches

Real-time Target Detection System in Scenic Landscape Based on Improved YOLOv4 Algorithm

Research on Data Enhancement Method for Complex Pavement

Contact Info

Product

Resources

About