Metric Learning On Cub 200 2011

评估指标

R@1

评测结果

各个模型在此基准测试上的表现结果

模型名称	R@1	Paper Title	Repository
PDDM Quadruplet	58.3	Local Similarity-Aware Deep Feature Embedding	-
ResNet50 + Language	71.4	Integrating Language Guidance into Vision-based Deep Metric Learning	-
EPSHN(64)	57.3	Improved Embeddings with Easy Positive Triplet Mining	-
BN-Inception + Proxy-Anchor	71.1	Proxy Anchor Loss for Deep Metric Learning	-
SCT(64)	57.7	Hard negative examples are hard, but useful	-
ResNet-50 + ProxyNCA++	69.0	ProxyNCA++: Revisiting and Revitalizing Proxy Neighborhood Component Analysis	-
BN-Inception + Group Loss	65.5	The Group Loss for Deep Metric Learning	-
ABE-8-512	60.6	Attention-based Ensemble for Deep Metric Learning	-
ABE + HORDE	66.8	Metric Learning With HORDE: High-Order Regularizer for Deep Embeddings	-
NED	74.9	Calibrated neighborhood aware confidence measure for deep metric learning	-
EPSHN(512)	64.9	Improved Embeddings with Easy Positive Triplet Mining	-
ResNet-50 + Cross-Entropy	69.2	A unifying mutual information view of metric learning: cross-entropy vs. pairwise losses	-
ResNet-50 + Intra-Batch Connections	71.8	Learning Intra-Batch Connections for Deep Metric Learning	-
EfficientDML-VPTSP-G/512	88.5	Learning Semantic Proxies from Visual Prompts for Parameter-Efficient Fine-Tuning in Deep Metric Learning	-
ResNet50 + NIR	70.5	Non-isotropy Regularization for Proxy-based Deep Metric Learning	-
CCL (ResNet-50)	73.45	Center Contrastive Loss for Metric Learning	-
MS + DIML	68.15	Towards Interpretable Deep Metric Learning with Structural Matching	-
ResNet50 (128) + PADS	67.3	PADS: Policy-Adapted Sampling for Visual Similarity Learning	-
Unicom+ViT-L@336px	90.1	Unicom: Universal and Compact Representation Learning for Image Retrieval	-
GoogLeNet + HDML	53.7	Hardness-Aware Deep Metric Learning	-

0 of 30 row(s) selected.