文献类型：期刊文献

英文题名：Few-shot image generation based on contrastive meta-learning generative adversarial network

作者：Phaphuangwittayakul, Aniwat[1];Ying, Fangli[2];Guo, Yi[1,3,4];Zhou, Liting[5];Chakpitak, Nopasit[6]

机构：[1]East China Univ Sci & Technol, Dept Comp Sci & Engn, Shanghai, Peoples R China;[2]East China Univ Sci & Technol, Dept Comp Sci & Engn, State Key Lab Bioreactor Engn, Shanghai, Peoples R China;[3]Natl Engn Lab Big Data Distribut & Exchange Techn, Shanghai, Peoples R China;[4]Shanghai Engn Res Ctr Big Data & Internet Audienc, Shanghai, Peoples R China;[5]Dublin City Univ, ADAPT Ctr, Sch Comp, Dublin 9, Ireland;[6]Chiang Mai Univ, Int Coll Digital Innovat, Chiang Mai, Thailand

年份：2023

卷号：39

期号：9

起止页码：4015

外文期刊名：VISUAL COMPUTER

收录：;EI(收录号：20223012405348);WOS:【SCI-EXPANDED(收录号:WOS:000828445600001)】；

基金：This research is financially supported by National Key Research and Development Program of China (No. 2020YFA0907800); is partially supported by open funding project of the State Key Laboratory of Bioreactor Engineering, East China University of Science and Technology, Shanghai, China; is also supported by The National Key Research and Development Program of China (Grant Number 2018YFC0807105) and Science and Technology Committee of Shanghai Municipality (STCSM) (under Grant Numbers 17DZ1101003, 18511106602 and 18DZ2252300); and International College of Digital Innovation (ICDI), Chiang Mai University, Thailand.

语种：英文

外文关键词：Few-shot image generation; Contrastive learning; Meta-learning; Generative adversarial network

摘要：Traditional deep generative models rely on enormous training data for generating images from a given class. However, they face the challenges associated with expensive and time-consuming in data acquisition as well as the requirements for fast learning from limited data of new categories. In this study, a contrastive meta-learning generative adversarial network (CML-GAN) is proposed to generate novel images of unseen classes from a few images by applying a self-supervised contrastive learning strategy to a fast adaptive meta-learning framework. By introducing a meta-learning framework into a GAN-based model, our model can efficiently learn the feature representations and quickly adapt to new generation tasks with only a few samples. The proposed model takes original input and generated images from the GAN-based model as inputs and evaluates both contrastive loss and distance loss based on the feature representations of the inputs extracted from the encoder. The original input image and its generated version from the generator are considered a positive pair, while the rest of the generated images in the same batch are considered negative samples. Then, the model converges to differentiate positive samples from negative ones and learns to generate distinct representations of the same samples, which prevents model overfitting. Thus, our model can generalize to generate diverse images from only a few samples of unseen categories, while fast adapting to new image generation tasks. Furthermore, the effectiveness of our model is demonstrated through extensive experiments on three datasets.