FAN-L-Hybrid (IN-22k)

Reported on 4 benchmarks across 2 tasks · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Methodology2 results

Domain AdaptationonImageNet-C
Top 1 Accuracy· uses extra data· 2022-04-26
73.6
SOTA
Understanding The Robustness in Vision Transformers arXiv:2204.12451
Domain AdaptationonImageNet-C
mean Corruption Error (mCE)· uses extra data· 2022-04-26
35.8
best: 22 (EfficientNet-L2+RPL)
Understanding The Robustness in Vision Transformers arXiv:2204.12451

Computer Vision2 results

Domain GeneralizationonImageNet-C
Top 1 Accuracy· uses extra data· 2022-04-26
73.6
SOTA
Understanding The Robustness in Vision Transformers arXiv:2204.12451
Domain GeneralizationonImageNet-C
mean Corruption Error (mCE)· uses extra data· 2022-04-26
35.8
best: 28.2 (DINOv2 (ViT-g/14, frozen model, linear eval))
Understanding The Robustness in Vision Transformers arXiv:2204.12451