Semi-Supervised Image Classification
Semi-supervised image classification leverages unlabelled data as well as labelled data to increase classification performance.
You may want to read some blog posts to get an overview before reading the papers and checking the leaderboards:
- An overview of proxy-label approaches for semi-supervised learning - Sebastian Ruder
- Semi-Supervised Learning in Computer Vision - Amit Chaudhary
( Image credit: Self-Supervised Semi-Supervised Learning )
Papers
Showing 1–10 of 167 papers
All datasetsImageNet - 10% labeled dataImageNet - 1% labeled dataCIFAR-10, 4000 Labelscifar-100, 10000 LabelsCIFAR-10, 250 LabelsSVHN, 1000 labelsSVHN, 250 LabelsCIFAR-10, 40 LabelsCIFAR-100, 400 LabelsCIFAR-10, 1000 LabelsCIFAR-100, 2500 LabelsSTL-10, 1000 Labels
Benchmark Results
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | MixMatch | Accuracy | 92.25 | — | Unverified |
| 2 | UPS (CNN-13) | Accuracy | 91.82 | — | Unverified |
| 3 | Triple-GAN-V2 (ResNet-26) | Accuracy | 91.59 | — | Unverified |
| 4 | LiDAM | Accuracy | 89.04 | — | Unverified |
| 5 | Dual Student (600) | Accuracy | 85.83 | — | Unverified |
| 6 | Triple-GAN-V2 (CNN-13) | Accuracy | 85 | — | Unverified |
| 7 | ICT (CNN-13) | Accuracy | 84.52 | — | Unverified |
| 8 | SESEMI SSL (ConvNet) | Accuracy | 82.12 | — | Unverified |
| 9 | Triple-GAN-V2 (CNN-13, no aug) | Accuracy | 81.81 | — | Unverified |