Long Tail Learning On Cifar 100 Lt R 100

評価指標

Error Rate

評価結果

このベンチマークにおける各モデルのパフォーマンス結果

モデル名	Error Rate	Paper Title	Repository
CE+DRS+GIT	55.65	Do Deep Networks Transfer Invariances Across Classes?
NCL(ResNet32)	46.7	Nested Collaborative Learning for Long-Tailed Visual Recognition	-
ELP	57.6	A Simple Episodic Linear Probe Improves Visual Recognition in the Wild
FBL (Resnet-32)	54.78	Feature-Balanced Loss for Long-Tailed Visual Recognition
Cross-Entropy (CE)	62.75	Revisiting Long-tailed Image Classification: Survey and Benchmarks with New Evaluation Metrics	-
LDAM-DRW + CMO	52.8	The Majority Can Help The Minority: Context-rich Minority Oversampling for Long-tailed Classification
BCL(ResNet-32)	46.1	Balanced Contrastive Learning for Long-Tailed Visual Recognition
LTR-weight-balancing	46.45	Long-Tailed Recognition via Weight Balancing
CBD+TailCalibX	53.41	Feature Generation for Long-tail Classification
MBJ	54.2	Memory-based Jitter: Improving Visual Recognition on Long-tailed Data with Diversity In Memory	-
PaCo + SAM	47.0	Escaping Saddle Points for Effective Generalization on Class-Imbalanced Data
RIDE + CMO + Curvature Regularization	49.3	Predicting and Enhancing the Fairness of DNNs with the Curvature of Perceptual Manifolds
GLMC+MaxNorm (ResNet-34, channel x4)	41.59	Global and Local Mixture Consistency Cumulative Learning for Long-tailed Visual Recognitions
LDAM-DRW-RSG	55.5	RSG: A Simple but Effective Module for Learning Imbalanced Datasets
OTLM+CE (Resnet-32)	53.90	Optimal Transport for Long-Tailed Recognition with Learnable Cost Matrix	-
CE-DRW	58.9	The Majority Can Help The Minority: Context-rich Minority Oversampling for Long-tailed Classification
GLMC + SAM	40.99	Escaping Saddle Points for Effective Generalization on Class-Imbalanced Data
VS + SAM	53.4	Escaping Saddle Points for Effective Generalization on Class-Imbalanced Data
TADE	50.2	Self-Supervised Aggregation of Diverse Experts for Test-Agnostic Long-Tailed Recognition	-
RIDE	52	Long-tailed Recognition by Routing Diverse Distribution-Aware Experts

0 of 66 row(s) selected.