论文链接:[2305.06144] Learning Semi-supervised Gaussian Mixture Models for Generalized Category Discovery (arxiv.org) 代码链接:github.com/DTennant/GPC 这篇文章中忽略了许多细节推导问题,事实上是引用了聚类中使用GMM的工作,需要理解细节的可以参考DeepDPM arxiv.org/abs/2203.1430。其实笔者也关注过DeepDPM,但...
给不同的class 产生单独对应的cluster就更完美了,后者相对于前者的简单的分类任务更困难,但最终得到的模型也更加鲁棒,因为当模型不仅能正确区分不同类别,并且可以给每个class产生独立的cluster(即不仅增大了类间距离,同时也减少了类内距离,从而使得整个representation spaces非常的紧凑而分明),那么这样的决策平面会更加rob...
Generalized Category Discovery (GCD) is a crucial task that aims to recognize both known and novel categories from a set of unlabeled data by utilizing a few labeled data with only known categories. Due to the lack of supervision and category information, current methods usually perform poorly on...
Otholt, J., Meinel, C., & Yang, H. (2024). Guided cluster aggregation: A hierarchical approach to generalized category discovery. InIEEE/CVF Winter conference on applications of computer vision (WACV), (pp. 2618–2627) Ouldnoughi, R., Kuo, C.-W., & Kira, Z. (2023). CLIP-GCD: ...
Generalized category discoveryThis work focuses on the problem of Generalized Category Discovery (GCD), a more realistic and challenging semi-supervised learning setting where unlabeled data may belong to either previously known or unseen categories. Recent advancements have demonstrated the efficacy of ...
Continuous Generalized Category Discovery (C-GCD) aims to continually discover novel classes from unlabelled image sets while maintaining performance on old classes. In this paper, we propose a novel learning framework, dubbed Neighborhood Commonality-aware Evolution Network (NCENet) that conquers this ...
GCDSS - Generalized Category Discovery in Semantic Segmentation Welcome to the official repository for the Generalized Category Discovery in Semantic Segmentation (GCDSS) project! Overview This repository contains the code implementation for GCDSS. We are currently in the process of organizing and refinin...
One major conundrum with both Generalized Category Discovery and Novel Category Discovery is that the definition of the category has remained undetermined. This complication can be overlooked when the granularity of categories at test time is similar to training time. However, for more realistic applica...
(2022a). Generalized category discovery. In CVPR. Vaze, S., Han, K., Vedaldi, A., & Zisserman, A. (2022b). Open-set recognition: A good closed-set classifier is all you need. In ICLR. Vernekar, S., Gaurav, A., Abdelzad, V., Denouden, T., Salay, R., & Czarnecki, K. ...
代码链接:DTennant/Incremental-Generalized-Category-Discovery (github.com) 摘要 我们探讨了增量通用类别发现(IGCD)问题。这是一个极具挑战性的增量分类学习问题,其目标是开发出能够正确分类以前所见类别图像的模型,同时发现新的类别。学习是在一系列时间步骤中进行的,模型在每次迭代时都会获得新的标记和未标记数据,并...