S4L: Self-Supervised Semi-Supervised Learning

This work tackles the problem of semi-supervised learning of image classifiers. Our main insight is that the field of semi-supervised learning can benefit from the quickly advancing field of self-supervised visual representation learning. Unifying these two approaches, we propose the framework of self-supervised semi-supervised learning and use it to derive two novel semi-supervised image classification methods. We demonstrate the effectiveness of these methods in comparison to both carefully tuned baselines, and existing semi-supervised learning methods. We then show that our approach and existing semi-supervised methods can be jointly trained, yielding a new state-of-the-art result on semi-supervised ILSVRC-2012 with 10% of labels.

PDF Abstract ICCV 2019 PDF ICCV 2019 Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Semi-Supervised Image Classification ImageNet - 10% labeled data Exemplar Fine-tuned (ResNet-50) Top 5 Accuracy 81.01% # 45
Semi-Supervised Image Classification ImageNet - 10% labeled data VAT + Entropy Minimization Top 5 Accuracy 83.39% # 39
Semi-Supervised Image Classification ImageNet - 10% labeled data Exemplar (joint training) Top 5 Accuracy 83.72% # 36
Semi-Supervised Image Classification ImageNet - 10% labeled data VAT (ResNet-50) Top 5 Accuracy 82.78% # 41
Semi-Supervised Image Classification ImageNet - 10% labeled data S4L-Exemplar (ResNet-50) Top 5 Accuracy 83.72% # 36
Semi-Supervised Image Classification ImageNet - 10% labeled data S4L-Rotation (ResNet-50) Top 5 Accuracy 83.82% # 33
Semi-Supervised Image Classification ImageNet - 10% labeled data Pseudolabeling (ResNet-50) Top 5 Accuracy 82.41% # 43
Semi-Supervised Image Classification ImageNet - 10% labeled data Rotation Fine-tuned (ResNet-50) Top 5 Accuracy 78.53% # 48
Semi-Supervised Image Classification ImageNet - 10% labeled data Rotation + VAT + Ent. Min. Top 5 Accuracy 91.23% # 15
Semi-Supervised Image Classification ImageNet - 10% labeled data S4L-MOAM (ResNet-50 4×) Top 5 Accuracy 91.23% # 15
Top 1 Accuracy 73.21% # 29
Semi-Supervised Image Classification ImageNet - 10% labeled data VAT + Entropy Minimization (ResNet-50) Top 5 Accuracy 83.39% # 39
Semi-Supervised Image Classification ImageNet - 10% labeled data Rotation Top 5 Accuracy 78.53% # 48
Semi-Supervised Image Classification ImageNet - 10% labeled data Exemplar Top 5 Accuracy 81.01% # 45
Semi-Supervised Image Classification ImageNet - 10% labeled data Pseudolabeling Top 5 Accuracy 82.41% # 43
Semi-Supervised Image Classification ImageNet - 10% labeled data VAT Top 5 Accuracy 82.78% # 41
Semi-Supervised Image Classification ImageNet - 1% labeled data Exemplar Top 5 Accuracy 44.90% # 38
Semi-Supervised Image Classification ImageNet - 1% labeled data Rotation Top 5 Accuracy 45.11% # 37
Semi-Supervised Image Classification ImageNet - 1% labeled data VAT + Entropy Minimization Top 5 Accuracy 46.96% # 36
Semi-Supervised Image Classification ImageNet - 1% labeled data Exemplar (joint training) Top 5 Accuracy 47.02% # 35
Semi-Supervised Image Classification ImageNet - 1% labeled data Pseudolabeling Top 5 Accuracy 51.56% # 34
Semi-Supervised Image Classification ImageNet - 1% labeled data Rotation (joint training) Top 5 Accuracy 53.37% # 33
Semi-Supervised Image Classification ImageNet - 1% labeled data VAT Top 5 Accuracy 44.05% # 39

Methods