arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2301.03976cs.CV
收藏

基于伪负标记的半监督学习的图像分类

Semi-Supervised Learning with Pseudo-Negative Labels for Image Classification

Ningbo University(宁波大学);Zhejiang Keyongtai Automation Technology Co., Ltd.(浙江科永泰自动化技术有限公司)

Hao Xu, Hui Xiao, Huazheng Hao, Li Dong, Xiaojie Qiu, Chengbin Peng

摘要:

半监督学习框架通常采用具有多个子模型的相互学习方法来从不同的角度学习。为了避免在这些子模型之间传输错误的伪标签,通常使用高阈值来过滤掉大量未标记数据的低置信度预测。然而,这种过滤不能充分利用预测置信度低的未标记数据。为了克服这个问题,在这项工作中,我们提出了一个基于伪负标签的相互学习框架。负标签是指对应的数据项不属于的标签。在每次迭代中,一个子模型为每个数据项生成伪负标签,另一个子模型从这些标签中学习。两个子模型的角色在每次迭代后交换,直到收敛。通过降低伪负标签上的预测概率,对偶模型可以提高其预测能力。我们还提出了一种机制来选择一些伪负标签来馈送到子模型中。在实验中,我们的框架在几个主要的基准测试中取得了最先进的结果。具体来说,使用我们的框架,13层CNN模型的错误率分别为9.35%和7.94%,CIFAR-10分别为1000和4000个标签。此外,对于只有20个标签的非增强MNIST,我们的框架的出错率为0.81%,这是远远小于其他方法。我们的方法也证明了一个显着的性能改善领域自适应。

英文摘要:

Semi-supervised learning frameworks usually adopt mutual learning approaches with multiple submodels to learn from different perspectives. To avoid transferring erroneous pseudo labels between these submodels, a high threshold is usually used to filter out a large number of low-confidence predictions for unlabeled data. However, such filtering can not fully exploit unlabeled data with low prediction confidence. To overcome this problem, in this work, we propose a mutual learning framework based on pseudo-negative labels. Negative labels are those that a corresponding data item does not belong. In each iteration, one submodel generates pseudo-negative labels for each data item, and the other submodel learns from these labels. The role of the two submodels exchanges after each iteration until convergence. By reducing the prediction probability on pseudo-negative labels, the dual model can improve its prediction ability. We also propose a mechanism to select a few pseudo-negative labels to feed into submodels. In the experiments, our framework achieves state-of-the-art results on several main benchmarks. Specifically, with our framework, the error rates of the 13-layer CNN model are 9.35% and 7.94% for CIFAR-10 with 1000 and 4000 labels, respectively. In addition, for the non-augmented MNIST with only 20 labels, the error rate is 0.81% by our framework, which is much smaller than that of other approaches. Our approach also demonstrates a significant performance improvement in domain adaptation.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑