文献类型：期刊文献

英文题名：Few-shot classification via efficient meta-learning with hybrid optimization

作者：Jia, Jinfang;Feng, Xiang[1];Yu, Huiqun

机构：[1]East China Univ Sci & Technol, Dept Comp Sci & Engn, Shanghai 200237, Peoples R China; Shanghai Engn Res Ctr Smart Energy, Shanghai 200237, Peoples R China

年份：2024

卷号：127

外文期刊名：ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE

收录：;EI(收录号：20234214912533);WOS:【SCI-EXPANDED(收录号:WOS:001096365500001)】；

基金：This work is supported by the National Natural Science Foundation of China (No. 62276097) , Key Program of National Natural Science Foundation of China (No. 62136003) , National Key Research and Development Program of China (No. 2020YFB1711700) , Special Fund for Information Development of Shanghai Economic and Information Commission, China (No. XX-XXFZ-02-20-2463) and Scientific Research Program of Shanghai Science and Technology Commission, China (No. 21002411000) .

语种：英文

外文关键词：Few-shot learning; Meta-learning; Data augmentation; Initialization attenuation; Resolution increase

摘要：Meta-learning is one of the important methods to solve the challenging few-shot learning setting by using previous knowledge and experience to guide the learning of new tasks. Model-agnostic meta-learning (MAML) is one of the most popular meta-learning algorithms, and many variants of MAML have appeared in recent years. However, the performance of this algorithm for few-shot classification falls behind some other algorithms working on this problem. Therefore, its generalization performance needs to be further explored and improved. In view of the generalization problem, we found that MAML always shares an initialization in the process of parameter update, ignoring the bias between different tasks, resulting in limited generalization performance. On the other hand, the sample diversity of meta-learning model is low, and shallow network training is generally used, so it is difficult to obtain good performance based on deep neural network models. Based on these problems, we propose a hybrid optimization meta-learning method based on data augmentation, initialization attenuation, and resolution increase, called Mix-MAML. Experimental results show that our method reaches 76.93% classification accuracy on mini-ImageNet with 100 x 100 resolution, and 83.62% classification accuracy on CIFAR-FS with 80 x 80 resolution in the 5-way 5-shot settings under ResNet12, which achieves comparable or even better performance than other algorithms in some standard few-shot learning benchmarks without changing MAML simplicity and model-agnostic.