Fine Grained Image Classification On Stanford

Metrics

Accuracy

PARAMS

Results

Performance results of various models on this benchmark

Model Name	Accuracy	PARAMS	Paper Title	Repository
DeiT-B	93.3%	86M	Training data-efficient image transformers & distillation through attention	-
SEB+EfficientNet-B5	94.6%	-	On the Eigenvalues of Global Covariance Pooling for Fine-grained Visual Recognition
S3N	94.7%	-	Selective Sparse Sampling for Fine-Grained Image Recognition
PMG	95.1%	-	Fine-Grained Visual Classification via Progressive Multi-Granularity Training of Jigsaw Patches
PCA	94.6%	-	Progressive Co-Attention Network for Fine-grained Visual Classification
SaSPA + CAL	95.72	-	Advancing Fine-Grained Classification by Structure and Subject Preserving Augmentation
ViT-L (attn finetune)	93.8%	-	Three things everyone should know about Vision Transformers
AENet	94.0%	-	Alignment Enhancement Network for Fine-grained Visual Categorization	-
ResMLP-12	84.6%	-	ResMLP: Feedforward networks for image classification with data-efficient training
SR-GNN	96.1	30.9	SR-GNN: Spatial Relation-aware Graph Neural Network for Fine-Grained Image Categorization
RDNet-T (224 res, IN-1K pretrained)	93.9%	24M	DenseNets Reloaded: Paradigm Shift Beyond ResNets and ViTs
ResNet50 (A1)	92.7%	24M	ResNet strikes back: An improved training procedure in timm
Inceptionv4	95.35%	-	Non-binary deep transfer learning for image classification
WS-DAN	94.5%	-	See Better Before Looking Closer: Weakly Supervised Data Augmentation Network for Fine-Grained Visual Classification
CCFR	95.5%	-	Re-rank Coarse Classification with Local Region Enhanced Features for Fine-Grained Image Recognition	-
MPSA	95.4%	-	Multi-Granularity Part Sampling Attention for Fine-Grained Visual Classification
CIN	94.5%	-	Channel Interaction Networks for Fine-Grained Image Categorization	-
DF-GMM	94.8%	-	Weakly Supervised Fine-Grained Image Classification via Guassian Mixture Model Oriented Discriminative Learning	-
MPFG + CLIP	86.79	-	Multiscale patch-based feature graphs for image classification
TransFG	94.8%	-	TransFG: A Transformer Architecture for Fine-grained Recognition

0 of 82 row(s) selected.