摘要
训练集和测试集之间的分布差距使深度学习模型在泛化方面面临挑战。通过系统分析,可发现2个亟待解决的关键问题:训练数据到测试数据的知识迁移优化不足,以及数据集中类别不均衡的影响。针对这些挑战,提出了一种新颖的测试时适应算法——可信伪标签微调方法(FTP)。通过优化样本选择过程,筛选出熵值较低的测试样本构建微调数据集,并结合原训练集实现模型微调,显著提高图像分类模型在测试集上的泛化性能。在MNIST,FashionMNIST和CIFAR10数据集上进行了广泛实验,结果表明结合FTP的图像分类模型在测试集上普遍获得性能提升,准确率最高提升约3%,F1分数相应提高,且优于TENT,COTTA,EATA和OSTTA等当前常用测试时适应方法。此外,基于梯度的可视化分析证明,经FTP微调的模型在保持高预测准确性的同时,依然维持了良好的可解释性,为实际应用提供了可靠保障。
The distribution gap between training and test sets poses generalization challenges for deep-learning models.Systematic analysis identified two key urgent problems:insufficient optimization of knowledge transfer from training data to test data,and the impact of uneven categories in the dataset.To address these challenges,a novel test-time adaptation algorithm,fine-tuning with trusted pseudo-labels(FTP),was proposed.By optimizing the sample selection process,the test samples with low entropy value were selected to construct a fine-tuned dataset,and the model was fine-tuned by combining with the original training set,significantly enhancing the generalization performance of the image classification model on the test set.Extensive experiments on MNIST,FashionMNIST,and CIFAR10 datasets showed that the image classification model combined with FTP generally achieved a performance improvement on the test set,with an accuracy increase of up to about 3%,and a corresponding increase in F1 score,outperforming the current commonly used test adaptation methods such as TENT,COTTA,EATA,and OSTTA.In addition,the gradient-based visual analysis confirmed that the FTP-fine-tuned model preserved good interpretability while maintaining high prediction accuracy,offering reliable guarantee for practical applications.
作者
李星辰
李宗民
杨超智
LI Xingchen;LI Zongmin;YANG Chaozhi(School of Computer Science and Technology,China University of Petroleum(East China),Qingdao Shandong 266580,China;Shengli College,China University of Petroleum,Dongying Shandong 257061,China)
出处
《图学学报》
北大核心
2025年第6期1292-1303,共12页
Journal of Graphics
基金
国家重点研发计划(2019YFF0301800)
国家自然科学基金(61379106)
山东省自然科学基金(ZR2013FM036,ZR2015FM011)。
关键词
深度学习
测试时适应
可信伪标签
模型微调
deep learning
test-time adaptation
trusted pseudo-labels
model fine-tuning
