HyperAI超神経

ホームニュース最新論文チュートリアルデータセット百科事典 SOTA LLMモデル GPU ランキング学会

サイトについて

日本語

HyperAI超神経

Image Classification On Inaturalist

評価指標

Top 1 Accuracy

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	Top 1 Accuracy	Paper Title	Repository
MetaSAug	63.28%	MetaSAug: Meta Semantic Augmentation for Long-Tailed Visual Recognition
AIMv2-1B	79.7	Multimodal Autoregressive Pre-training of Large Vision Encoders
SpineNet-143	63.6%	SpineNet: Learning Scale-Permuted Backbone for Recognition and Localization
iSQRT-COV-Net	-	Deep CNNs Meet Global Covariance Pooling: Better Representation and Generalization
MAE (ViT-H, 448)	83.4	Masked Autoencoders Are Scalable Vision Learners
b_22DeiT-LT(ours)	-	DeiT-LT Distillation Strikes Back for Vision Transformer Training on Long-Tailed Datasets
AIMv2-H	77.9	Multimodal Autoregressive Pre-training of Large Vision Encoders
Hiera-H (448px)	83.8	Hiera: A Hierarchical Vision Transformer without the Bells-and-Whistles
FixSENet-154	75.4	Fixing the train-test resolution discrepancy
IncResNetV2 SE	67.3%	The iNaturalist Species Classification and Detection Dataset
MetaFormer (MetaFormer-2,384,extra_info)	83.4%	MetaFormer: A Unified Meta Framework for Fine-Grained Recognition
MetaFormer (MetaFormer-2,384)	80.4%	MetaFormer: A Unified Meta Framework for Fine-Grained Recognition
TransFG	71.7	TransFG: A Transformer Architecture for Fine-grained Recognition
Graph-RISE (40M)	31.12%	Graph-RISE: Graph-Regularized Image Semantic Embedding
SEB+EfficientNet-B5	72.3	On the Eigenvalues of Global Covariance Pooling for Fine-grained Visual Recognition
AIMv2-3B	81.5	Multimodal Autoregressive Pre-training of Large Vision Encoders
AIMv2-L	76	Multimodal Autoregressive Pre-training of Large Vision Encoders
AIMv2-3B (448 res)	85.9	Multimodal Autoregressive Pre-training of Large Vision Encoders

0 of 18 row(s) selected.