SOTAVerified

Self-Supervised Learning

Self-Supervised Learning is proposed for utilizing unlabeled data with the success of supervised learning. Producing a dataset with good labels is expensive, while unlabeled data is being generated all the time. The motivation of Self-Supervised Learning is to make use of the large amount of unlabeled data. The main idea of Self-Supervised Learning is to generate the labels from unlabeled data, according to the structure or characteristics of the data itself, and then train on this unsupervised data in a supervised manner. Self-Supervised Learning is wildly used in representation learning to make a model learn the latent features of the data. This technique is often employed in computer vision, video processing and robot control.

Source: Self-supervised Point Set Local Descriptors for Point Cloud Registration

Image source: LeCun

Papers

Showing 10011050 of 5044 papers

TitleStatusHype
Unsupervised Domain Generalization by Learning a Bridge Across DomainsCode1
3rd Place: A Global and Local Dual Retrieval Solution to Facebook AI Image Similarity ChallengeCode1
Self-Supervised Material and Texture Representation Learning for Remote Sensing TasksCode1
Novel Class Discovery in Semantic SegmentationCode1
The Surprising Effectiveness of Representation Learning for Visual ImitationCode1
TransFGU: A Top-down Approach to Fine-Grained Unsupervised Semantic SegmentationCode1
A Fast Knowledge Distillation Framework for Visual RecognitionCode1
D2C: Diffusion-Decoding Models for Few-Shot Conditional GenerationCode1
TTT++: When Does Self-Supervised Test-Time Training Fail or Thrive?Code1
Container: Context Aggregation NetworksCode1
Object-Aware Cropping for Self-Supervised LearningCode1
Self-supervised Autoregressive Domain Adaptation for Time Series DataCode1
Similarity Contrastive Estimation for Self-Supervised Soft Contrastive LearningCode1
Self-Supervised Pre-Training of Swin Transformers for 3D Medical Image AnalysisCode1
SLA^2P: Self-supervised Anomaly Detection with Adversarial PerturbationCode1
Robust Equivariant Imaging: a fully unsupervised framework for learning to image from noisy and partial measurementsCode1
DABS: A Domain-Agnostic Benchmark for Self-Supervised LearningCode1
Self-Supervised Pre-Training for Transformer-Based Person Re-IdentificationCode1
Music Classification: Beyond Supervised Learning, Towards Real-world ApplicationsCode1
SSR: An Efficient and Robust Framework for Learning with Unknown Label NoiseCode1
Why Do Self-Supervised Models Transfer? Investigating the Impact of Invariance on Downstream TasksCode1
Benchmarking Detection Transfer Learning with Vision TransformersCode1
Self-supervised Re-renderable Facial Albedo Reconstruction from Single ImageCode1
2nd Place Solution to Facebook AI Image Similarity Challenge Matching TrackCode1
Masked Autoencoders Are Scalable Vision LearnersCode1
Self-Supervised Multi-Object Tracking with Cross-Input ConsistencyCode1
Self-Supervised Audio-Visual Representation Learning with Relaxed Cross-Modal SynchronicityCode1
Do we still need ImageNet pre-training in remote sensing scene classification?Code1
Generalized Radiograph Representation Learning via Cross-supervision between Images and Free-text Radiology ReportsCode1
Towards the Generalization of Contrastive Self-Supervised LearningCode1
Learning Continuous Representation of Audio for Arbitrary Scale Super ResolutionCode1
Contrastive prediction strategies for unsupervised segmentation and categorization of phonemes and wordsCode1
ReSkin: versatile, replaceable, lasting tactile skinsCode1
Self-Supervised Learning Disentangled Group Representation as FeatureCode1
Equivariant Contrastive LearningCode1
OMASGAN: Out-of-Distribution Minimum Anomaly Score GAN for Sample Generation on the BoundaryCode1
Hyper-Representations: Self-Supervised Representation Learning on Neural Network Weights for Model Characteristic PredictionCode1
Intermediate Layers Matter in Momentum Contrastive Self Supervised LearningCode1
Towards artificial general intelligence via a multimodal foundation modelCode1
Self-supervised EEG Representation Learning for Automatic Sleep StagingCode1
Robust Contrastive Learning Using Negative Samples with Diminished SemanticsCode1
Self-supervised similarity search for large scientific datasetsCode1
Contrastive Neural Processes for Self-Supervised LearningCode1
Distance-wise Prototypical Graph Neural Network in Node Imbalance ClassificationCode1
A Simple Baseline for Low-Budget Active LearningCode1
CLOOB: Modern Hopfield Networks with InfoLOOB Outperform CLIPCode1
Understanding Dimensional Collapse in Contrastive Self-supervised LearningCode1
SentimentArcs: A Novel Method for Self-Supervised Sentiment Analysis of Time Series Shows SOTA Transformers Can Struggle Finding Narrative ArcsCode1
Mitigating Memorization of Noisy Labels via Regularization between RepresentationsCode1
TLDR: Twin Learning for Dimensionality ReductionCode1
Show:102550
← PrevPage 21 of 101Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1Pretraining: NoneImages & Text57.5Unverified
2Pretraining: ShEDImages & Text54.3Unverified
3Pretraining: e-MixImages & Text48.9Unverified
#ModelMetricClaimedVerifiedStatus
1ResNet50Accuracy91.7Unverified
2ResNet18Accuracy91.02Unverified
3MV-MRAccuracy89.67Unverified
#ModelMetricClaimedVerifiedStatus
1ResNet50average top-1 classification accuracy93.89Unverified
2ResNet18average top-1 classification accuracy92.58Unverified
#ModelMetricClaimedVerifiedStatus
1ResNet50average top-1 classification accuracy72.51Unverified
2ResNet18average top-1 classification accuracy69.31Unverified
#ModelMetricClaimedVerifiedStatus
1CorInfomax (ResNet50)Top-1 Accuracy82.64Unverified
2CorInfomax (ResNet18)Top-1 Accuracy80.48Unverified
#ModelMetricClaimedVerifiedStatus
1ResNet50average top-1 classification accuracy51.84Unverified
2ResNet18average top-1 classification accuracy51.67Unverified
#ModelMetricClaimedVerifiedStatus
1CorInfomax (ResNet18)Top-1 Accuracy93.18Unverified
#ModelMetricClaimedVerifiedStatus
1CorInfomax (ResNet18)Top-1 Accuracy71.61Unverified
#ModelMetricClaimedVerifiedStatus
1Hybrid BYOL-S/CvTAccuracy67.2Unverified
#ModelMetricClaimedVerifiedStatus
1CorInfomax (ResNet50)Top-1 Accuracy54.86Unverified