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 2210.04402cs.CV
收藏

深度度量学习的对比贝叶斯分析

Contrastive Bayesian Analysis for Deep Metric Learning

Central South University(中南大学);Beijing Jiaotong University(北京交通大学);Shenzhen University(深圳大学);University of Missouri(密苏里大学);University of Kragujevac(克拉古耶瓦茨大学);Southern University of Science and Technology(南方科技大学);Pengcheng Lab(鹏城实验室)

Shichao Kan, Zhiquan He, Yigang Cen, Yang Li, Vladimir Mladenovic, Zhihai He

摘要:

最近的深度度量学习方法一直专注于在正样本和负样本对之间设计不同的对比损失函数,以便学习的特征嵌入能够将同一类的正样本拉得更近,并将不同类的负样本推离彼此。在这项工作中,我们认识到,有一个显着的语义差距之间的功能在中间的功能层和类标签在最终的输出层。为了弥合这一差距,我们开发了一个对比贝叶斯分析的特征和模型的后验概率的图像标签的条件下,他们的特征相似性在对比学习设置。这种对比贝叶斯分析为深度度量学习带来了一种新的损失函数。为了提高所提出的方法到新的类的泛化能力,我们进一步扩展了度量方差约束的对比贝叶斯损失。我们的实验结果和消融研究表明,所提出的对比贝叶斯度量学习方法显着提高了监督和伪监督场景下深度度量学习的性能,大大优于现有方法。

英文摘要:

Recent methods for deep metric learning have been focusing on designing different contrastive loss functions between positive and negative pairs of samples so that the learned feature embedding is able to pull positive samples of the same class closer and push negative samples from different classes away from each other. In this work, we recognize that there is a significant semantic gap between features at the intermediate feature layer and class labels at the final output layer. To bridge this gap, we develop a contrastive Bayesian analysis to characterize and model the posterior probabilities of image labels conditioned by their features similarity in a contrastive learning setting. This contrastive Bayesian analysis leads to a new loss function for deep metric learning. To improve the generalization capability of the proposed method onto new classes, we further extend the contrastive Bayesian loss with a metric variance constraint. Our experimental results and ablation studies demonstrate that the proposed contrastive Bayesian metric learning method significantly improves the performance of deep metric learning in both supervised and pseudo-supervised scenarios, outperforming existing methods by a large margin.

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。

↑