A Few-Shot Learning Model Based on Semi-Supervised with Pseudo Label

YU You; FENG Lin; WANG Ge-ge; XU Qi-feng

doi:10.3969/j.issn.0372-2112.2019.11.007

您当前的位置：

首页 >

文章列表页 >

A Few-Shot Learning Model Based on Semi-Supervised with Pseudo Label

更新时间：2025-07-16

- A Few-Shot Learning Model Based on Semi-Supervised with Pseudo Label
- Acta Electronica Sinica Vol. 47, Issue 11, Pages: 2284-2291(2019)
- 作者机构：
  
  四川师范大学计算机科学学院,四川,成都,610101
- 作者简介：
- 基金信息：
- DOI：10.3969/j.issn.0372-2112.2019.11.007
  CLC： TP181
- Published Online：25 November 2019，
  
  Published：2019
- 稿件说明：
移动端阅览
YU You, FENG Lin, WANG Ge-ge, et al. A Few-Shot Learning Model Based on Semi-Supervised with Pseudo Label[J]. Acta Electronica Sinica, 2019, 47(11): 2284-2291.
DOI：

YU You, FENG Lin, WANG Ge-ge, et al. A Few-Shot Learning Model Based on Semi-Supervised with Pseudo Label[J]. Acta Electronica Sinica, 2019, 47(11): 2284-2291. DOI： 10.3969/j.issn.0372-2112.2019.11.007.

摘要

如何将带有大量标记数据的源域知识模型迁移至带有少量标记数据的目标域是少样本学习研究领域的热点问题.针对现有的少样本学习算法在源域数据与目标域数据的特征分布差异较大时存在的泛化能力较弱的问题，提出一种基于伪标签的半监督少样本学习模型FSLSS（Few-Shot Learning based on Semi-Supervised）.首先，利用pytorch深度学习框架建立一个关系型深度学习网络，并使用源域数据对网络进行预训练；然后，使用此网络对目标域数据进行分类预测，将分类概率最大的类标签作为数据的伪标签；最后，利用目标域的伪标签数据和源域的真实标签数据对网络进行混合训练，并重复伪标签标记与混合训练过程.实验结果表明，相对于现有主流少样本学习算法，FSLSS模型有更好的泛化能力及知识迁移效果.

Abstract

How to migrate a source domain knowledge model with a large amount of tagged data to a target domain with a small amount of tagged data is a hot issue in few-shot learning. For the problems that the existing few-shot learning algorithm have weak generalization ability when the difference between the feature distribution of the source domain data and the target domain data is large

a few-shot learning model based on semi-supervised FSLSS (Few-Shot Learning based on Semi-Supervised) is proposed. Firstly

a relational deep learning network is established by using the pytorch framework

and the network is pre-trained by the source domain data. Then

the network is used to predict the target domain data

and the label with the highest classification probability is used as the data's pseudo label. Finally

the network is hybrid trained using the pseudo label data of the target domain and the real label data of the source domain

then repeating the pseudo-labeled and hybrid trained process. The experimental results show that the FSLSS model has better generalization ability and knowledge transfer effect than the existing few-shot learning algorithms.

关键词

Keywords

references

Views

237

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Locomotion Intention Prediction via Multi-Sensor Fusion and Transfer Learning for Lower Limb Exoskeletons

Hybrid Task Assignment for Crowd Sensing Based on Mobility Prediction

Semi-Supervised Medical Image Segmentation Based on Suspicious Pixel Mutual Correction

Semantic Comparison Model for Binary Programs Based on Statistical Reasoning

Local Representation Few-Shot Learning for Hyperspectral Image Cross-Scene Classification

Related Author

YU Zhi-peng

WANG Mei-ling

WANG Cheng-jun

LING Liu-yi

JIN Li

JIANG Wei-jin

NIE Cai-yan

LIU Qian

Related Institution

School of Artificial Intelligence, Anhui University of Science Technology

State Key Laboratory of Digital Intelligent Technology for Unmanned Coal Mining, Anhui University of Science Technology

Anhui Key Laboratory of Bionic Sensing and Advanced Robot technology

Institute of Intelligent Machinery, Hefei Institutes of Physical Science, Chinese Academy of Sciences

School of Computer Science, Hunan University of Technology and Business

⁰