Generalized Category Discovery

Vaze, Sagar; Hant, Kai; Vedaldi, Andrea; Zisserman, Andrew

doi:10.1109/cvpr52688.2022.00734

Cited by 128 publications

(113 citation statements)

References 24 publications

Supporting

Mentioning

110

Contrasting

Order By: Relevance

“…In [12,38], a k-means is performed on the entire dataset D l ∪ D u . The number of unknown classes C u is estimated to be the k that maximized the Hungarian clustering accuracy (see Section 2): a k too high will result in clusters assigned to the null set and a number too low will have clusters composed of multiple classes, both cases will be considered as being assigned incorrectly.…”

Section: Estimating the Number Of Unknown Classesmentioning

confidence: 99%

“…Generalized Category Discovery (GCD) [12] is a setting that is gaining traction from the community, with some very recent articles published [12,39,41,38]. GCD was designed to be a less constrained and more realistic setting of Novel Class Discovery, as it does not assume that samples…”

Section: New Domains Derived From Novel Class Discoverymentioning

confidence: 99%

“…It is assumed that the model does not need to be able to distinguish the known from the unknown classes. If this assumption is not made, this becomes a Generalized Category Discovery (GCD) [12] problem. Some solutions have been proposed for the NCD problem in the context of computer vision and have displayed promising results [13,14,15,16].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Novel Class Discovery: an Introduction and Key Concepts

Troisemaine¹,

Lemaire²,

Gosselin³

et al. 2023

Preprint

View full text Add to dashboard Cite

Novel Class Discovery (NCD) is a growing field where we are given during training a labeled set of known classes and an unlabeled set of different classes that must be discovered. In recent years, many methods have been proposed to address this problem, and the field has begun to mature. In this paper, we provide a comprehensive survey of the state-of-the-art NCD methods. We start by formally defining the NCD problem and introducing important notions. We then give an overview of the different families of approaches, organized by the way they transfer knowledge from the labeled set to the unlabeled set. We find that they either learn in two stages, by first extracting knowledge from the labeled data only and then applying it to the unlabeled data, or in one stage by conjointly learning on both sets. For each family, we describe their general principle and detail a few representative methods. Then, we briefly introduce some new related tasks inspired by the increasing number of NCD works. We also present some common tools and techniques used in NCD, such as pseudo labeling, self-supervised learning and contrastive learning. Finally, to help readers unfamiliar with the NCD problem differentiate it from other closely related domains, we summarize some of the closest areas of research and discuss their main differences.

show abstract

Section: Estimating the Number Of Unknown Classesmentioning

confidence: 99%

Section: New Domains Derived From Novel Class Discoverymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Novel Class Discovery: an Introduction and Key Concepts

Troisemaine¹,

Lemaire²,

Gosselin³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Several works [4,47,55] showed the use of these features for various applications such as instance segmentation, object discovery or transfer learning on downstream tasks. GCD [52] categorize all unlabeled images given only a partially labeled dataset. Other works used the pre-trained features for both object localization and segmentation [29], and even object part discovery and segmentation [7].…”

Section: Dino-vit Features As Local Semantic Descriptorsmentioning

confidence: 99%

Neural Congealing: Aligning Images to a Joint Semantic Atlas

Ofri-Amar¹,

Geyer²,

Kasten³

et al. 2023

Preprint

View full text Add to dashboard Cite

Project webpage: https://neural-congealing.github.io/ Input Images Congealed Images Edited ImagesFigure 1. Given a set of input images, our method automatically detects and jointly aligns semantically-common content across the images. This is achieved through a test-time training approach that estimates a unified 2D atlas that represents the common semantic content, and dense mappings from the joint atlas to each of the input images. Our atlas and mappings are optimized per input set in a self-supervised manner by leveraging a pre-trained DINO-ViT model. Our method can be applied to diverse image sets, without requiring any additional training data, and allows us to automatically propagate an edit applied to a single image across the entire set.

show abstract

“…Most of these methods also assume the unlabeled data only contains instances from new classes or assume the information that whether an unlabeled instance is from new classes is known. The concurrent work by Vaze et al [53] extends NCD to a generalized setting where the unlabeled instances may come from both old and new classes, while still requiring access to labeled and unlabeled instances jointly. In contrast, with Novel Class Discovery without Forgetting, we introduce a staged learning and account for the performance on both labeled and the unlabeled data, without requiring access to the labeled data when learning on unlabeled data to discover new classes.…”

Section: Novel Class Discoverymentioning

confidence: 99%

Novel Class Discovery Without Forgetting

Joseph

Paul

Aggarwal

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Humans possess an innate ability to identify and differentiate instances that they are not familiar with, by leveraging and adapting the knowledge that they have acquired so far. Importantly, they achieve this without deteriorating the performance on their earlier learning. Inspired by this, we identify and formulate a new, pragmatic problem setting of NCDwF: Novel Class Discovery without Forgetting, which tasks a machine learning model to incrementally discover novel categories of instances from unlabeled data, while maintaining its performance on the previously seen categories. We propose 1) a method to generate pseudolatent representations which act as a proxy for (no longer available) labeled data, thereby alleviating forgetting, 2) a mutual-information based regularizer which enhances unsupervised discovery of novel classes, and 3) a simple Known Class Identifier which aids generalized inference when the testing data contains instances form both seen and unseen categories. We introduce experimental protocols based on CIFAR-10, CIFAR-100 and ImageNet-1000 to measure the trade-off between knowledge retention and novel class discovery. Our extensive evaluations reveal that existing models catastrophically forget previously seen categories while identifying novel categories, while our method is able to effectively balance between the competing objectives. We hope our work will attract further research into this newly identified pragmatic problem setting.

show abstract

Generalized Category Discovery

Cited by 128 publications

References 24 publications

Novel Class Discovery: an Introduction and Key Concepts

Novel Class Discovery: an Introduction and Key Concepts

Neural Congealing: Aligning Images to a Joint Semantic Atlas

Novel Class Discovery Without Forgetting

Contact Info

Product

Resources

About