Long Tail Learning On Cifar 10 Lt R 50

Error Rate

Résultats

Résultats de performance de divers modèles sur ce benchmark

Tableau comparatif

Nom du modèle	Error Rate
escaping-saddle-points-for-effective	8.44
deit-lt-distillation-strikes-back-for-vision	10.2
nested-collaborative-learning-for-long-tailed	13.2
metasaug-meta-semantic-augmentation-for-long	15.66
mdcs-more-diverse-experts-with-consistency	11.7
pure-noise-to-the-rescue-of-insufficient-data	10.8
enhanced-long-tailed-recognition-with	12.0
sure-survey-recipes-for-building-reliable-and	9.78