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 2201.09308cs.CV
收藏

基于篮子的Softmax

Basket-based Softmax

University of Chinese Academy of Sciences(中国科学院大学)

Qiang Meng, Xinqian Gu, Xiaqing Xu, Feng Zhou

摘要:

基于Softmax的损失在人脸识别和重新识别等各种任务上取得了最先进的性能。然而,这些方法高度依赖于具有全局标签的干净数据集,这限制了它们在许多现实应用中的使用。一个重要的原因是,合并和组织来自各种时间和空间场景的数据集通常是不现实的,因为可能会引入噪声标签,并且需要指数增加的资源。为了解决这个问题,我们提出了一种称为Basket-based Softmax(BBS)的新型训练期间挖掘策略及其并行版本,以端到端的方式在多个数据集上有效地训练模型。具体来说,对于每个训练样本,我们同时采用相似性得分作为从其他数据集中挖掘否定类的线索,并动态添加它们以辅助区分性特征的学习。在实验中,我们证明了BBS的效率和优越性的人脸识别和重新识别的任务,与模拟和现实世界的数据集。

英文摘要:

Softmax-based losses have achieved state-of-the-art performances on various tasks such as face recognition and re-identification. However, these methods highly relied on clean datasets with global labels, which limits their usage in many real-world applications. An important reason is that merging and organizing datasets from various temporal and spatial scenarios is usually not realistic, as noisy labels can be introduced and exponential-increasing resources are required. To address this issue, we propose a novel mining-during-training strategy called Basket-based Softmax (BBS) as well as its parallel version to effectively train models on multiple datasets in an end-to-end fashion. Specifically, for each training sample, we simultaneously adopt similarity scores as the clue to mining negative classes from other datasets, and dynamically add them to assist the learning of discriminative features. Experimentally, we demonstrate the efficiency and superiority of the BBS on the tasks of face recognition and re-identification, with both simulated and real-world datasets.

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。

↑