Combined Hyperbolic and Euclidean Soft Triple Loss Beyond the Single Space Deep Metric Learning

Open paper

Benchmark	Model	Rank	Results
metric-learning-on-cars196	CHEST-ViT (ViT-B-E, 1024)	–	R@1: 91.9
metric-learning-on-cars196	CHEST-ViT (ViT-B-H, 512)	–	R@1: 93.1
metric-learning-on-cars196	CHEST-ViT (ViT-S-E, 384)	–	R@1: 88.5
metric-learning-on-cars196	CHEST-ViT (ViT-S-H, 384)	–	R@1: 89.1
metric-learning-on-cub-200-2011	CHEST-ViT (ViT-B-E, 1024)	–	R@1: 88.7
metric-learning-on-cub-200-2011	CHEST-ViT (ViT-B-H, 512)	–	R@1: 88.8
metric-learning-on-cub-200-2011	CHEST-ViT (ViT-S-E, 384)	–	R@1: 86.0
metric-learning-on-cub-200-2011	CHEST-ViT (ViT-S-H, 384)	–	R@1: 86.6