Image Classification On Omnibenchmark

评估指标

Average Top-1 Accuracy

评测结果

各个模型在此基准测试上的表现结果

		Paper Title
NOAH-ViTB/16	47.6	Neural Prompt Search
SwinTransformer	46.4	Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
Bamboo-R50	45.4	Bamboo: Building Mega-Scale Vision Dataset Continually with Human-Machine Synergy
Adapter-ViTB/16	44.5	Parameter-Efficient Transfer Learning for NLP
CLIP-RN50	42.1	Learning Transferable Visual Models From Natural Language Supervision
IG-1B	40.4	Billion-scale semi-supervised learning for image classification
BiT-M	40.4	Big Transfer (BiT): General Visual Representation Learning
DINO	38.9	Emerging Properties in Self-Supervised Vision Transformers
SwAV	38.3	Unsupervised Learning of Visual Features by Contrasting Cluster Assignments
ResNet-101	37.4	Deep Residual Learning for Image Recognition
MEAL-V2	36.6	MEAL V2: Boosting Vanilla ResNet-50 to 80%+ Top-1 Accuracy on ImageNet without Tricks
MoPro-V2	36.1	MoPro: Webly Supervised Learning with Momentum Prototypes
EfficientNetB4	35.8	EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
MoCoV2	34.8	Momentum Contrast for Unsupervised Visual Representation Learning
ResNet-50	34.3	Deep Residual Learning for Image Recognition
InceptionV4	32.3	Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning
MLP-Mixer	32.2	MLP-Mixer: An all-MLP Architecture for Vision
Manifold	31.6	Manifold Mixup: Better Representations by Interpolating Hidden States
CutMix	31.1	CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features
ReLabel	30.8	Re-labeling ImageNet: from Single to Multi-Labels, from Global to Localized Labels

0 of 22 row(s) selected.