Pretraining: e-Mix

Reported on 6 benchmarks across 1 task · 1 paper · 2 SOTA

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

Computer Vision6 results

Self-Supervised LearningonDABS
Natural Images· 2021-11-23
27.9
SOTA
DABS: A Domain-Agnostic Benchmark for Self-Supervised Learning arXiv:2111.12062
Self-Supervised LearningonDABS
Speech· 2021-11-23
41.8
SOTA
DABS: A Domain-Agnostic Benchmark for Self-Supervised Learning arXiv:2111.12062
Self-Supervised LearningonDABS
Images & Text· 2021-11-23
48.9
best: 57.5 (Pretraining: None)
DABS: A Domain-Agnostic Benchmark for Self-Supervised Learning arXiv:2111.12062
Self-Supervised LearningonDABS
Med. Imaging· 2021-11-23
72.4
best: 74.5 (Pretraining: ShED)
DABS: A Domain-Agnostic Benchmark for Self-Supervised Learning arXiv:2111.12062
Self-Supervised LearningonDABS
Sensors· 2021-11-23
79.5
best: 88.7 (Pretraining: ShED)
DABS: A Domain-Agnostic Benchmark for Self-Supervised Learning arXiv:2111.12062
Self-Supervised LearningonDABS
Text· 2021-11-23
44.1
best: 48.4 (Pretraining: ShED)
DABS: A Domain-Agnostic Benchmark for Self-Supervised Learning arXiv:2111.12062