Multi-scale conditional reconstruction generative adversarial network

doi:10.1016/j.imavis.2023.104885

	Multi-scale conditional reconstruction generative adversarial network
	Chen, Yanming 1; Xu, Jiahao 1; An, Zhulin 2; Zhuang, Fuzhen 3
	2024
发表期刊	IMAGE AND VISION COMPUTING
ISSN	0262-8856
卷号	141 页码:9
摘要	Generative adversarial network has become the factual standard for high-quality image synthesis. However, modeling the distribution of complex datasets (e.g. ImageNet and COCO-Stuff) remains challenging in unsupervised approaches. This is partly due to the imbalance between the generator and the discriminator during training, the discriminator easily defeats the generator because of special views. In this paper, we propose a model called multi-scale conditional reconstruction GAN (MS-GAN). The core concept of MS-GAN is to model the local density implicitly using different scales of instance conditions. Instance conditions are extracted from the target images via a self-supervised learning model. In addition, we alignment the semantic features of the observed instances by adding an additional reconstruction loss to the generator. Our MS-GAN can aggregate instance features at different scales and maximize semantic features. This allows the generator to learn additional comparative knowledge from instance features, leading to a better feature representation, thus improving the performance of the generation task. Experimental results on the ImageNet dataset and the COCO-Stuff dataset show that our method matches or exceeds the original performance in both FID and IS scores compared to the ICGAN framework. Additionally, our precision score on the ImageNet dataset improved from 74.2% to 79.9%.
关键词	Generative adversarial network Unsupervised generation Multi-scale instance Reconstructed losses
DOI	10.1016/j.imavis.2023.104885
收录类别	SCI
语种	英语
资助项目	National Science Foundation of China (NSFC)[62262067] ; Key Natural Science Foundation of Education Department of Anhui[KJ2021A0046]
WOS研究方向	Computer Science ; Engineering ; Optics
WOS类目	Computer Science, Artificial Intelligence ; Computer Science, Software Engineering ; Computer Science, Theory & Methods ; Engineering, Electrical & Electronic ; Optics
WOS记录号	WOS:001145154200001
出版者	ELSEVIER
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/38412
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	An, Zhulin
作者单位	1.Anhui Univ, Sch Compute Sci & Technol, Hefei, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Beijing, Peoples R China 3.Beihang Univ, Inst Artificial Intelligence, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Chen, Yanming,Xu, Jiahao,An, Zhulin,et al. Multi-scale conditional reconstruction generative adversarial network[J]. IMAGE AND VISION COMPUTING,2024,141:9.
APA	Chen, Yanming,Xu, Jiahao,An, Zhulin,&Zhuang, Fuzhen.(2024).Multi-scale conditional reconstruction generative adversarial network.IMAGE AND VISION COMPUTING,141,9.
MLA	Chen, Yanming,et al."Multi-scale conditional reconstruction generative adversarial network".IMAGE AND VISION COMPUTING 141(2024):9.