arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.02662cs.CV
收藏

基于语义对比学习的深度聚类

Deep Clustering by Semantic Contrastive Learning

Queen Mary University of London(伦敦玛丽女王大学)

Jiabo Huang, Shaogang Gong

摘要:

虽然对比学习最近通过学习样本特定的判别性视觉特征为未标记图像的深度聚类带来了显着的好处,但其显式推断类决策边界的潜力还不太清楚。这是因为其实例判别策略不是类敏感的,因此,在所得样本特定特征空间上导出的聚类未被优化以对应于有意义的类决策边界。在这项工作中,我们解决了这个问题,通过引入语义对比学习(SCL)。SCL通过制定语义(聚类感知)对比学习目标,在未标记的训练数据上明确地施加基于距离的聚类结构。此外,我们引入了一个聚类的一致性条件,以满足共同的实例视觉相似性和集群决策边界,并同时优化两者的原因有关的假设,语义地面真相类(未知/未标记)的飞行由他们的共识。这种发现未知类决策边界的语义对比学习方法对于对象识别任务的无监督学习具有相当大的优势。大量的实验表明,SCL在六个对象识别基准测试上优于最先进的对比学习和深度聚类方法,特别是在更具挑战性的细粒度和较大的数据集上。

英文摘要:

Whilst contrastive learning has recently brought notable benefits to deep clustering of unlabelled images by learning sample-specific discriminative visual features, its potential for explicitly inferring class decision boundaries is less well understood. This is because its instance discrimination strategy is not class sensitive, therefore, the clusters derived on the resulting sample-specific feature space are not optimised for corresponding to meaningful class decision boundaries. In this work, we solve this problem by introducing Semantic Contrastive Learning (SCL). SCL imposes explicitly distance-based cluster structures on unlabelled training data by formulating a semantic (cluster-aware) contrastive learning objective. Moreover, we introduce a clustering consistency condition to be satisfied jointly by both instance visual similarities and cluster decision boundaries, and concurrently optimising both to reason about the hypotheses of semantic ground-truth classes (unknown/unlabelled) on-the-fly by their consensus. This semantic contrastive learning approach to discovering unknown class decision boundaries has considerable advantages to unsupervised learning of object recognition tasks. Extensive experiments show that SCL outperforms state-of-the-art contrastive learning and deep clustering methods on six object recognition benchmarks, especially on the more challenging finer-grained and larger 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。

↑