VALSE 论文速览 第200期：基于亲和邻域优化的深度聚类算法

论文题目：

Contextually Neighborhood Refinery for Deep Clustering

作者列表：

Chunlin Yu, Ye Shi, and Jingya Wang

B站观看网址：

https://www.bilibili.com/video/BV1voxNe8E8w/?spm_id_from=333.999.0.0&vd_source=2044ac986b5caa4c8b5f00b525441df3

复制链接到浏览器打开或点击阅读原文即可跳转至观看页面。

论文摘要：

Previous endeavors in self-supervised learning have enlightened the research of deep clustering from an instance discrimination perspective. Built upon this foundation, recent studies further highlight the importance of grouping semantically similar instances. One effective method to achieve this is by promoting the semantic structure preserved by neighborhood consistency. However, the samples in the local neighborhood may be limited due to their close proximity to each other, which may not provide substantial and diverse supervision signals. Inspired by the versatile re-ranking methods in the context of image retrieval, we propose to employ an efficient online re-ranking process to mine more informative neighbors in a Contextually Affinitive (ConAff) Neighborhood, and then encourage the cross-view neighborhood consistency. To further mitigate the intrinsic neighborhood noises near cluster boundaries, we propose a progressively relaxed boundary filtering strategy to circumvent the issues brought by noisy neighbors. Our method can be easily integrated into the generic self-supervised frameworks and outperforms the state-of-the-art methods on several popular benchmarks. 

论文链接：

[https://arxiv.org/abs/2312.07806]

代码链接：

[https://github.com/cly234/DeepClustering-ConNR]

视频讲者简介：

余春霖，上海科技大学信息学院研究生，研究方向为计算机视觉，主要研究行人再识别、增量学习、自监督学习。

个人主页：

https://cly234.github.io/

特别鸣谢本次论文速览主要组织者：

月度轮值AC：于茜 (北京航空航天大学)