paper-with-me

홈 › Papers

Self-Supervised Learning by Estimating Twin Class Distributions

2021-10-14 · Feng Wang, Tao Kong, Rufeng Zhang, Huaping Liu, Hang Li

We present TWIST, a simple and theoretically explainable self-supervised representation learning method by classifying large-scale unlabeled datasets in an end-to-end way. We employ a siamese network terminated by a softmax operation to produce twin class distributions of two augmented images. Without supervision, we enforce the class distributions of different augmentations to be consistent. However, simply minimizing the divergence between augmentations will cause collapsed solutions, i.e., outputting the same class probability distribution for all images. In this case, no information about the input image is left. To solve this problem, we propose to maximize the mutual information between the input and the class predictions. Specifically, we minimize the entropy of the distribution for each sample to make the class prediction for each sample assertive and maximize the entropy of the mean distribution to make the predictions of different samples diverse. In this way, TWIST can naturally avoid the collapsed solutions without specific designs such as asymmetric network, stop-gradient operation, or momentum encoder. As a result, TWIST outperforms state-of-the-art methods on a wide range of tasks. Especially, TWIST performs surprisingly well on semi-supervised learning, achieving 61.2% top-1 accuracy with 1% ImageNet labels using a ResNet-50 as backbone, surpassing previous best results by an absolute improvement of 6.2%. Codes and pre-trained models are given on: https://github.com/bytedance/TWIST

📄 PDF Abstract BibTeX arXiv:2110.07402

Code (2)

bytedance/TWIST 공식 구현 pytorch
beresandras/contrastive-classification-keras tf

Tasks

Fine-Grained Image ClassificationImage ClassificationRepresentation LearningSelf-Supervised Image ClassificationSelf-Supervised LearningSemi-Supervised Image ClassificationTransfer LearningUnsupervised Image Classification

Methods 이 논문이 사용한 방법론

Siamese Network 설명 없음
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…

Similar Papers 제목 키워드 기반

Audio Barlow Twins: Self-Supervised Audio Representation Learning

2022-09-28 · Jonah Anton, Harry Coppock, Pancham Shukla, Bjorn W. Schuller

The Barlow Twins self-supervised learning objective requires neither negative samples or asymmetric learning updates, achieving results on a par with the current state-of-the-art within Computer Vision. As such, we prese…

Environmental Sound ClassificationEvent DetectionRepresentation LearningSelf-Supervised Learning

Barlow Twins: Self-Supervised Learning via Redundancy Reduction

2021-03-04 · Jure Zbontar, Li Jing, Ishan Misra, Yann Lecun 외

Self-supervised learning (SSL) is rapidly closing the gap with supervised methods on large computer vision benchmarks. A successful approach to SSL is to learn embeddings which are invariant to distortions of the input s…

General ClassificationImage ClassificationObject DetectionSelf-Supervised Image Classification+2

Infinite Width Limits of Self Supervised Neural Networks

2024-11-17 · Maximilian Fleissner, Gautham Govind Anil, Debarghya Ghoshdastidar

The NTK is a widely used tool in the theoretical analysis of deep learning, allowing us to look at supervised deep neural networks through the lenses of kernel regression. Recently, several works have investigated kernel…

Self-Supervised Learning

Self-supervised Knowledge Distillation for Few-shot Learning

2020-06-17 · Jathushan Rajasegaran, Salman Khan, Munawar Hayat, Fahad Shahbaz Khan 외

Real-world contains an overwhelmingly large number of object classes, learning all of which at once is infeasible. Few shot learning is a promising learning paradigm due to its ability to learn out of order distributions…

Few-Shot Image ClassificationFew-Shot LearningKnowledge DistillationMeta-Learning+1

A Histopathology Study Comparing Contrastive Semi-Supervised and Fully Supervised Learning

2021-11-10 · Lantian Zhang, Mohamed Amgad, Lee A. D. Cooper

Data labeling is often the most challenging task when developing computational pathology models. Pathologist participation is necessary to generate accurate labels, and the limitations on pathologist time and demand for …

Active LearningSelf-Supervised LearningWeakly-supervised Learning