Characterizing and Improving the Robustness of Self-Supervised Learning through Background Augmentations
Open paper
Benchmark
Model
Rank
Results
image-classification-on-objectnet
BYOL (BG_RM)
–
Top-1 Accuracy: 23.9
image-classification-on-objectnet
MoCo-v2 (BG_Swaps)
–
Top-1 Accuracy: 20.8
image-classification-on-objectnet
SwAV (BG_RM)
–
Top-1 Accuracy: 21.9
Rank counts only results with a code link.