<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>In the era of aesthetic economy, users pay more and more attention to the aesthetic quality and emotional experience of products. The study discusses the intelligent design method of product model integrating computational aesthetics, combines the advantages of variational self-encoder and generative adversarial network, constructs a generative model based on VAE-GAN, and quantitatively analyzes the product images generated by the VAE-GAN network to explore the effect of the products it generates. Then, take the car form generated by the model as an example, establish the visual presentation quantitative index system of the product model form layout, and evaluate the visual presentation effect of the generated car model after applying gray correlation analysis to assign the indexes. The VAE-GAN model has a better generative effect, and its SSIM value is greater than 0.8 in most of the training phases. Among them, the product model generated after 2000 steps of training is the most excellent, with the highest evaluation score. The product model is the most outstanding, with the highest evaluation score of 1.92. The overall visual presentation effect of the product forms generated by the model is good, and the comprehensive evaluation scores are above 0.6. The VAE-GAN model can be used for product generation, and the visual presentation evaluation method that integrates computational aesthetics can realize the aesthetic evaluation of product forms, which can help designers make design creations and program decisions.</p>
</abstract>

In the era of aesthetic economy, users pay more and more attention to the aesthetic quality and emotional experience of products. The study discusses the intelligent design method of product model integrating computational aesthetics, combines the advantages of variational self-encoder and generative adversarial network, constructs a generative model based on VAE-GAN, and quantitatively analyzes the product images generated by the VAE-GAN network to explore the effect of the products it generates. Then, take the car form generated by the model as an example, establish the visual presentation quantitative index system of the product model form layout, and evaluate the visual presentation effect of the generated car model after applying gray correlation analysis to assign the indexes. The VAE-GAN model has a better generative effect, and its SSIM value is greater than 0.8 in most of the training phases. Among them, the product model generated after 2000 steps of training is the most excellent, with the highest evaluation score. The product model is the most outstanding, with the highest evaluation score of 1.92. The overall visual presentation effect of the product forms generated by the model is good, and the comprehensive evaluation scores are above 0.6. The VAE-GAN model can be used for product generation, and the visual presentation evaluation method that integrates computational aesthetics can realize the aesthetic evaluation of product forms, which can help designers make design creations and program decisions.

Research on Product Model Generation and Visual Presentation Incorporating Computational Aesthetics

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

In the era of aesthetic economy, users pay more and more attention to the aesthetic quality and emotional experience of products. The study discusses the...

Research on Product Model Generation and Visual Presentation Incorporating Computational Aesthetics

Online veröffentlicht: 05. Aug. 2024

Seitenbereich: -

Eingereicht: 19. Apr. 2024

Akzeptiert: 03. Juli 2024

DOI: https://doi.org/10.2478/amns-2024-2240

Schlüsselwörter
Generative adversarial networks, VAE-GAN, Gray correlation analysis, Visual presentation effects, Computational aesthetics

© 2024 Xinjia Zhang, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Research on Product Model Generation and Visual Presentation Incorporating Computational Aesthetics

Online veröffentlicht: 05. Aug. 2024

Seitenbereich: -

Eingereicht: 19. Apr. 2024

Akzeptiert: 03. Juli 2024

DOI: https://doi.org/10.2478/amns-2024-2240

SchlüsselwörterGenerative adversarial networks, VAE-GAN, Gray correlation analysis, Visual presentation effects, Computational aesthetics

© 2024 Xinjia Zhang, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Schlüsselwörter
Generative adversarial networks, VAE-GAN, Gray correlation analysis, Visual presentation effects, Computational aesthetics