arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.04455cs.CV
收藏

生物特征识别的连续表示学习

Continual Representation Learning for Biometric Identification

The University of Edinburgh(爱丁堡大学);SenseTime Group Limited(商汤科技);The University of Sydney(悉尼大学)

Bo Zhao, Shixiang Tang, Dapeng Chen, Hakan Bilen, Rui Zhao

摘要:

随着近年来数字数据的爆炸式增长,从数据流中不断学习新任务而不忘记以前获得的知识变得越来越重要。在本文中,我们提出了一个新的持续学习(CL)设置,即"连续表示学习“,其重点是学习更好的表示在一个连续的方式。我们还提供了两个大规模的多步骤生物识别基准,其中不同类别的视觉外观是高度相关的。与要求模型识别更多的学习类相反,我们的目标是学习特征表示,这些特征表示不仅可以更好地推广到以前看不见的图像,还可以推广到看不见的类/身份。对于新的设置,我们提出了一种新的方法,执行知识蒸馏在大量的身份通过应用邻域选择和一致性松弛策略,以提高可扩展性和灵活性的持续学习模型。我们证明,现有的CL方法可以提高在新的设置的代表性,我们的方法取得了更好的结果比竞争对手。

英文摘要:

With the explosion of digital data in recent years, continuously learning new tasks from a stream of data without forgetting previously acquired knowledge has become increasingly important. In this paper, we propose a new continual learning (CL) setting, namely ``continual representation learning'', which focuses on learning better representation in a continuous way. We also provide two large-scale multi-step benchmarks for biometric identification, where the visual appearance of different classes are highly relevant. In contrast to requiring the model to recognize more learned classes, we aim to learn feature representation that can be better generalized to not only previously unseen images but also unseen classes/identities. For the new setting, we propose a novel approach that performs the knowledge distillation over a large number of identities by applying the neighbourhood selection and consistency relaxation strategies to improve scalability and flexibility of the continual learning model. We demonstrate that existing CL methods can improve the representation in the new setting, and our method achieves better results than the competitors.

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。

↑