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 2010.04592cs.LGstat.ML
收藏

具有硬负样本的对比学习

Contrastive Learning with Hard Negative Samples

Massachusetts Institute of Technology(麻省理工学院)

Joshua Robinson, Ching-Yao Chuang, Suvrit Sra, Stefanie Jegelka

摘要:

你如何能为对比学习挑选好的反面例子?我们认为,与度量学习一样,表征的对比学习受益于硬负样本(即,难以与锚点区分的点)。使用硬阴性的关键挑战是对比方法必须保持无监督,使得采用使用真实相似性信息的现有阴性采样策略不可行。作为回应,我们开发了一个新的家庭的无监督采样方法选择硬负样品,用户可以控制硬度。这种抽样的一个限制情况导致了一种表示,这种表示紧密地聚集了每个类,并将不同的类尽可能地分开。所提出的方法提高了跨多种模态的下游性能,只需要很少的额外代码行来实现,并且不引入计算开销。

英文摘要:

How can you sample good negative examples for contrastive learning? We argue that, as with metric learning, contrastive learning of representations benefits from hard negative samples (i.e., points that are difficult to distinguish from an anchor point). The key challenge toward using hard negatives is that contrastive methods must remain unsupervised, making it infeasible to adopt existing negative sampling strategies that use true similarity information. In response, we develop a new family of unsupervised sampling methods for selecting hard negative samples where the user can control the hardness. A limiting case of this sampling results in a representation that tightly clusters each class, and pushes different classes as far apart as possible. The proposed method improves downstream performance across multiple modalities, requires only few additional lines of code to implement, and introduces no computational overhead.

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。

↑