paper-with-me

Papers

Are all negatives created equal in contrastive instance discrimination?

2020-10-13 · Tiffany Tianhui Cai, Jonathan Frankle, David J. Schwab, Ari S. Morcos

Self-supervised learning has recently begun to rival supervised learning on computer vision tasks. Many of the recent approaches have been based on contrastive instance discrimination (CID), in which the network is trained to recognize two augmented versions of the same instance (a query and positive) while discriminating against a pool of other instances (negatives). The learned representation is then used on downstream tasks such as image classification. Using methodology from MoCo v2 (Chen et al., 2020), we divided negatives by their difficulty for a given query and studied which difficulty ranges were most important for learning useful representations. We found a minority of negatives -- the hardest 5% -- were both necessary and sufficient for the downstream task to reach nearly full accuracy. Conversely, the easiest 95% of negatives were unnecessary and insufficient. Moreover, the very hardest 0.1% of negatives were unnecessary and sometimes detrimental. Finally, we studied the properties of negatives that affect their hardness, and found that hard negatives were more semantically similar to the query, and that some negatives were more consistently easy or hard than we would expect by chance. Together, our results indicate that negatives vary in importance and that CID may benefit from more intelligent negative treatment.

📄 PDF Abstract BibTeX arXiv:2010.06682

Code (0)

등록된 구현이 없습니다.

Tasks

Allimage-classificationImage ClassificationSelf-Supervised Learning

Methods 이 논문이 사용한 방법론

Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Random Gaussian Blur Random Gaussian Blur is an image data augmentation technique where we randomly blur the image using a Gaussian distribution. Image Source:…
Feedforward Network A Feedforward Network, or a Multilayer Perceptron (MLP), is a neural network with solely densely connected layers. This is the classic neural network architecture of the…
Batch Normalization 설명 없음
MoCo 설명 없음
InfoNCE 설명 없음
MoCo v2 MoCo v2 is an improved version of the Momentum Contrast self-supervised learning algorithm. Motivated by the findings presented in…

Similar Papers 제목 키워드 기반

Time-Series Contrastive Learning against False Negatives and Class Imbalance

2023-12-19 · Xiyuan Jin, Jing Wang, Lei Liu, Youfang Lin

As an exemplary self-supervised approach for representation learning, time-series contrastive learning has exhibited remarkable advancements in contemporary research. While recent contrastive learning strategies have foc…

Contrastive LearningRepresentation LearningTime Series

Robust Audio-Visual Instance Discrimination

2021-03-29 · CVPR 2021 1 · Pedro Morgado, Ishan Misra, Nuno Vasconcelos

We present a self-supervised learning method to learn audio and video representations. Prior work uses the natural correspondence between audio and video to define a standard cross-modal instance discrimination task, whe…

Action RecognitionContrastive LearningSelf-Supervised LearningTransfer Learning

Beyond Instance Discrimination: Relation-aware Contrastive Self-supervised Learning

2022-11-02 · Yifei Zhang, Chang Liu, Yu Zhou, Weiping Wang 외

Contrastive self-supervised learning (CSL) based on instance discrimination typically attracts positive samples while repelling negatives to learn representations with pre-defined binary self-supervision. However, vanill…

RelationSelf-Supervised Learning

Whitening for Self-Supervised Representation Learning

2020-07-13 · Aleksandr Ermolov, Aliaksandr Siarohin, Enver Sangineto, Nicu Sebe

Most of the current self-supervised representation learning (SSL) methods are based on the contrastive loss and the instance-discrimination task, where augmented versions of the same image instance ("positives") are cont…

Representation LearningSelf-Supervised Learning

Unsupervised Visual Representation Learning by Synchronous Momentum Grouping

2022-07-13 · Bo Pang, Yifan Zhang, Yaoyi Li, Jia Cai 외

In this paper, we propose a genuine group-level contrastive visual representation learning method whose linear evaluation performance on ImageNet surpasses the vanilla supervised learning. Two mainstream unsupervised lea…

ClusteringContrastive LearningLinear evaluationRepresentation Learning+1