DLME (ResNet-50, linear)

Reported on 2 benchmarks across 1 task · 1 paper · 1 SOTA

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

Computer Vision2 results

Image ClassificationonImageNet-100 (TEMI Split)
Percentage correct· 2022-07-07
79.3
best: 86.96 (SparseSwin with L2)
SOTA
DLME: Deep Local-flatness Manifold Embedding arXiv:2207.03160
Image ClassificationonSTL-10
Percentage correct· 2022-07-07
90.1
best: 99.64 (µ2Net+ (ViT-L/16))
DLME: Deep Local-flatness Manifold Embedding arXiv:2207.03160