Image Classification On Inaturalist 2018

评估指标

Top-1 Accuracy

评测结果

各个模型在此基准测试上的表现结果

模型名称	Top-1 Accuracy	Paper Title	Repository
µ2Net+ (ViT-L/16)	80.97	A Continual Development Methodology for Large-scale Multitask Dynamic ML Systems
ResNet-50	49.7%	ClusterFit: Improving Generalization of Visual Representations
Barlow Twins (ResNet-50)	46.5	Barlow Twins: Self-Supervised Learning via Redundancy Reduction
LeViT-384	66.9%	LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference
ResNet-50	69.8%	Grafit: Learning fine-grained image representations with coarse labels	-
BS-CMO (ResNet-50)	74.0%	The Majority Can Help The Minority: Context-rich Minority Oversampling for Long-tailed Classification
CaiT-M-36 U 224	78%	Going deeper with Image Transformers
ResNeXt-101 (SAMix)	70.54%	Boosting Discriminative Visual Representation Learning with Scenario-Agnostic Mixup
GPaCo (ResNet-152)	78.1%	Generalized Parametric Contrastive Learning
RIDE (ResNet-50)	72.2%	Long-tailed Recognition by Routing Diverse Distribution-Aware Experts
ResNet-152	69.05%	Class-Balanced Loss Based on Effective Number of Samples
CeiT-T (384 finetune resolution)	72.2%	Incorporating Convolution Designs into Visual Transformers
CeiT-S (384 finetune resolution)	79.4%	Incorporating Convolution Designs into Visual Transformers
LeViT-128S	55.2%	LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference
RegNet-8GF	81.2%	Grafit: Learning fine-grained image representations with coarse labels	-
ResNet-50 (AutoMix)	64.73%	AutoMix: Unveiling the Power of Mixup for Stronger Classifiers
Hiera-H (448px)	87.3%	Hiera: A Hierarchical Vision Transformer without the Bells-and-Whistles
SWAG (ViT H/14)	86.0%	Revisiting Weakly Supervised Pre-Training of Visual Perception Models
LeViT-256	66.2%	LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference
LeViT-192	60.4%	LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference

0 of 60 row(s) selected.