Recently, many AI-aided layout design systems are developed to reduce tedious manual intervention based on deep learning. However, most methods focus on a specific generation task. This paper explores a challenging problem to obtain multiple layout design generation (LDG), which generates floor plan or urban plan from a boundary input under a unified framework. One of the main challenges of multiple LDG is to obtain reasonable topological structures of layout generation with irregular boundaries and layout elements for different types of design. This paper formulates the multiple LDG task as an image-to-image translation problem, and proposes a conditional generative adversarial network (GAN), called LDGAN, with adaptive modules. The framework of LDGAN is based on a generator-discriminator architecture, where the generator is integrated with conditional convolution constrained by the boundary input and the attention module with channel and spatial features. Qualitative and quantitative experiments were conducted on the SCUT-AutoALP and RPLAN datasets, and the comparison with the state-of-the-art methods illustrate the effectiveness and superiority of the proposed LDGAN.
Xing ZHU
South China University of Technology
Yuxuan LIU
Waseda University
Lingyu LIANG
South China University of Technology, Southeast University,Pazhou Lab
Tao WANG
Minjiang University,Wuyi University
Zuoyong LI
Minjiang University
Qiaoming DENG
South China University of Technology
Yubo LIU
South China University of Technology
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Xing ZHU, Yuxuan LIU, Lingyu LIANG, Tao WANG, Zuoyong LI, Qiaoming DENG, Yubo LIU, "Multiple Layout Design Generation via a GAN-Based Method with Conditional Convolution and Attention" in IEICE TRANSACTIONS on Information,
vol. E106-D, no. 9, pp. 1615-1619, September 2023, doi: 10.1587/transinf.2022EDL8106.
Abstract: Recently, many AI-aided layout design systems are developed to reduce tedious manual intervention based on deep learning. However, most methods focus on a specific generation task. This paper explores a challenging problem to obtain multiple layout design generation (LDG), which generates floor plan or urban plan from a boundary input under a unified framework. One of the main challenges of multiple LDG is to obtain reasonable topological structures of layout generation with irregular boundaries and layout elements for different types of design. This paper formulates the multiple LDG task as an image-to-image translation problem, and proposes a conditional generative adversarial network (GAN), called LDGAN, with adaptive modules. The framework of LDGAN is based on a generator-discriminator architecture, where the generator is integrated with conditional convolution constrained by the boundary input and the attention module with channel and spatial features. Qualitative and quantitative experiments were conducted on the SCUT-AutoALP and RPLAN datasets, and the comparison with the state-of-the-art methods illustrate the effectiveness and superiority of the proposed LDGAN.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2022EDL8106/_p
Copy
@ARTICLE{e106-d_9_1615,
author={Xing ZHU, Yuxuan LIU, Lingyu LIANG, Tao WANG, Zuoyong LI, Qiaoming DENG, Yubo LIU, },
journal={IEICE TRANSACTIONS on Information},
title={Multiple Layout Design Generation via a GAN-Based Method with Conditional Convolution and Attention},
year={2023},
volume={E106-D},
number={9},
pages={1615-1619},
abstract={Recently, many AI-aided layout design systems are developed to reduce tedious manual intervention based on deep learning. However, most methods focus on a specific generation task. This paper explores a challenging problem to obtain multiple layout design generation (LDG), which generates floor plan or urban plan from a boundary input under a unified framework. One of the main challenges of multiple LDG is to obtain reasonable topological structures of layout generation with irregular boundaries and layout elements for different types of design. This paper formulates the multiple LDG task as an image-to-image translation problem, and proposes a conditional generative adversarial network (GAN), called LDGAN, with adaptive modules. The framework of LDGAN is based on a generator-discriminator architecture, where the generator is integrated with conditional convolution constrained by the boundary input and the attention module with channel and spatial features. Qualitative and quantitative experiments were conducted on the SCUT-AutoALP and RPLAN datasets, and the comparison with the state-of-the-art methods illustrate the effectiveness and superiority of the proposed LDGAN.},
keywords={},
doi={10.1587/transinf.2022EDL8106},
ISSN={1745-1361},
month={September},}
Copy
TY - JOUR
TI - Multiple Layout Design Generation via a GAN-Based Method with Conditional Convolution and Attention
T2 - IEICE TRANSACTIONS on Information
SP - 1615
EP - 1619
AU - Xing ZHU
AU - Yuxuan LIU
AU - Lingyu LIANG
AU - Tao WANG
AU - Zuoyong LI
AU - Qiaoming DENG
AU - Yubo LIU
PY - 2023
DO - 10.1587/transinf.2022EDL8106
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E106-D
IS - 9
JA - IEICE TRANSACTIONS on Information
Y1 - September 2023
AB - Recently, many AI-aided layout design systems are developed to reduce tedious manual intervention based on deep learning. However, most methods focus on a specific generation task. This paper explores a challenging problem to obtain multiple layout design generation (LDG), which generates floor plan or urban plan from a boundary input under a unified framework. One of the main challenges of multiple LDG is to obtain reasonable topological structures of layout generation with irregular boundaries and layout elements for different types of design. This paper formulates the multiple LDG task as an image-to-image translation problem, and proposes a conditional generative adversarial network (GAN), called LDGAN, with adaptive modules. The framework of LDGAN is based on a generator-discriminator architecture, where the generator is integrated with conditional convolution constrained by the boundary input and the attention module with channel and spatial features. Qualitative and quantitative experiments were conducted on the SCUT-AutoALP and RPLAN datasets, and the comparison with the state-of-the-art methods illustrate the effectiveness and superiority of the proposed LDGAN.
ER -