Image Classification On Cifar 100

评估指标

Percentage correct

评测结果

各个模型在此基准测试上的表现结果

		Paper Title	Repository
EffNet-L2 (SAM)	96.08	Sharpness-Aware Minimization for Efficiently Improving Generalization
Swin-L + ML-Decoder	95.1	ML-Decoder: Scalable and Versatile Classification Head
µ2Net (ViT-L/16)	94.95	An Evolutionary Approach to Dynamic Introduction of Tasks in Large-scale Multitask Learning Systems
ViT-B-16 (ImageNet-21K-P pretrain)	94.2	ImageNet-21K Pretraining for the Masses
CvT-W24	94.09	CvT: Introducing Convolutions to Vision Transformers
ViT-B/16 (PUGD)	93.95	Perturbated Gradients Updating within Unit Space for Deep Learning
Heinsen Routing + BEiT-large 16 224	93.8	An Algorithm for Routing Vectors in Sequences
BiT-L (ResNet)	93.51	Big Transfer (BiT): General Visual Representation Learning
Astroformer	93.36	Astroformer: More Data Might not be all you need for Classification
VIT-L/16 (Spinal FC, Background)	93.31	Reduction of Class Activation Uncertainty with Background Information
CaiT-M-36 U 224	93.1	-	-
ViT-L (attn fine-tune)	93.0	Three things everyone should know about Vision Transformers
TResNet-L-V2	92.6	TResNet: High Performance GPU-Dedicated Architecture
EfficientNetV2-L	92.3	EfficientNetV2: Smaller Models and Faster Training
EfficientNetV2-M	92.2	EfficientNetV2: Smaller Models and Faster Training
BiT-M (ResNet)	92.17	Big Transfer (BiT): General Visual Representation Learning
CeiT-S (384 finetune resolution)	91.8	Incorporating Convolution Designs into Visual Transformers
CeiT-S	91.8	Incorporating Convolution Designs into Visual Transformers
EfficientNet-B7	91.7	EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
EfficientNetV2-S	91.5	EfficientNetV2: Smaller Models and Faster Training

0 of 210 row(s) selected.