Few-shot classification on VGG Flower Meta-Dataset 1.0 (test)
76.21Accuracy (1-shot)SOSN + ArL
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| SOSN + ArLLearning Protocol=Supervised2020.01 | 76.21 | — | — | 88.36 | |
| SoSN - RL(att.)Learning Protocol=Supervised2020.01 | 74.96 | — | — | 87.21 | |
| SOSN - AL(att.)Learning Protocol=Supervised2020.01 | 74.31 | — | — | 86.97 | |
| SOSN - AL(cls.)Learning Protocol=Supervised2020.01 | 72.97 | — | — | 85.35 | |
| SOSNLearning Protocol=Supervised2020.01 | 71.9 | — | — | 84.87 | |
| SOSN - RL(cls.)Learning Protocol=Supervised2020.01 | 71.9 | — | — | 84.87 | |
| RN + ArLLearning Protocol=Supervised2020.01 | 71.12 | — | — | 83.49 | |
| U-SOSN + ArLLearning Protocol=Unsupervised2020.01 | 69.14 | — | — | 84.1 | |
| RNLearning Protocol=Supervised2020.01 | 68.26 | — | — | 80.94 | |
| PNLearning Protocol=Supervised2020.01 | 62.81 | — | — | 82.11 | |
| U-SOSNLearning Protocol=Unsupervised2020.01 | 61.26 | — | — | 75.98 | |
| U-PN + ArLLearning Protocol=Unsupervised2020.01 | 57.61 | — | — | 70.31 | |
| U-RN + ArLLearning Protocol=Unsupervised2020.01 | 57.19 | — | — | 70.23 | |
| U-RNLearning Protocol=Unsupervised2020.01 | 55.54 | — | — | 68.86 | |
| U-PNLearning Protocol=Unsupervised2020.01 | 55.36 | — | — | 68.49 | |
| BIGAN(knn)Learning Protocol=Unsupervised2020.01 | 44.68 | — | — | 59.12 | |
| FinetuneTraining source=all datasets2019.03 | — | 85.28 | 3 | — | |
| fo-MAMLTraining source=all datasets2019.03 | — | 79.94 | 6 | — | |
| k-NNTraining source=all datasets2019.03 | — | 83.47 | 4 | — | |
| MatchingNetTraining source=all datasets2019.03 | — | 81.9 | 5 | — | |
| Proto-MAMLTraining source=all datasets2019.03 | — | 88.72 | 1 | — | |
| ProtoNetTraining source=all datasets2019.03 | — | 86.85 | 2 | — | |
| RelationNetTraining source=all datasets2019.03 | — | 76.08 | 7 | — |