Next Article in Journal
Valorisation of Sugar Beet Processing Waste for Enhanced Production of Fungal Biomass as a Sustainable Alternative Protein Source
Previous Article in Journal
Effects of Hip-Dominant Strength Exercises on Post-Activation Performance Enhancement
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Applied Sciences
Volume 15
Issue 24
10.3390/app152413163
applsci-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
Article

Semantic-Aware 3D GAN: CLIP-Guided Disentanglement for Efficient Cross-Category Shape Generation

by
Weinan Cai
1,2,3,
Zongji Wang
1,2,
Yuanben Zhang
1,2,*,
Zhihong Zeng
1,2,
Xinming Li
1,2,3 and
Junyi Liu
1,2,3
1
Aerospace Information Research Institute of Chinese Academy of Sciences, Beijing 100019, China
2
Key Laboratory of Target Cognition and Application Technology (TCAT), Beijing 100019, China
3
The School of Electronic, University of Chinese Academy of Sciences, Beijing 100019, China
*
Author to whom correspondence should be addressed.
Appl. Sci. 2025, 15(24), 13163; https://doi.org/10.3390/app152413163
Submission received: 9 November 2025 / Revised: 5 December 2025 / Accepted: 11 December 2025 / Published: 15 December 2025
(This article belongs to the Section Computing and Artificial Intelligence)
Review Reports Versions Notes

Abstract

Generative Adversarial Networks (GANs) have achieved remarkable success in image generation. Although GAN-based approaches have also advanced three-dimensional (3D) data synthesis, they exhibit stagnation when compared to other state-of-the-art 3D generative models. Current 3D GAN methods suffer from training efficiency, generation diversity, and generalization in their original architectures. Among those challenges, cross-category training and generation are especially important in causing the degradation of synthesized results. In this paper, we propose a novel 3D generation framework to explore the capability boundaries of 3D GANs. The method features a novel style-based mechanism for controlling shape generation, a corresponding training procedure, and a CLIP-guided joint optimization scheme. This approach effectively mitigates generation diversity issues while maintaining generation quality and training stability.
Keywords: deep learning; 3D generation; generative adversarial networks deep learning; 3D generation; generative adversarial networks

Share and Cite

MDPI and ACS Style

Cai, W.; Wang, Z.; Zhang, Y.; Zeng, Z.; Li, X.; Liu, J. Semantic-Aware 3D GAN: CLIP-Guided Disentanglement for Efficient Cross-Category Shape Generation. Appl. Sci. 2025, 15, 13163. https://doi.org/10.3390/app152413163

AMA Style

Cai W, Wang Z, Zhang Y, Zeng Z, Li X, Liu J. Semantic-Aware 3D GAN: CLIP-Guided Disentanglement for Efficient Cross-Category Shape Generation. Applied Sciences. 2025; 15(24):13163. https://doi.org/10.3390/app152413163

Chicago/Turabian Style

Cai, Weinan, Zongji Wang, Yuanben Zhang, Zhihong Zeng, Xinming Li, and Junyi Liu. 2025. "Semantic-Aware 3D GAN: CLIP-Guided Disentanglement for Efficient Cross-Category Shape Generation" Applied Sciences 15, no. 24: 13163. https://doi.org/10.3390/app152413163

APA Style

Cai, W., Wang, Z., Zhang, Y., Zeng, Z., Li, X., & Liu, J. (2025). Semantic-Aware 3D GAN: CLIP-Guided Disentanglement for Efficient Cross-Category Shape Generation. Applied Sciences, 15(24), 13163. https://doi.org/10.3390/app152413163

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop