HER2 reader: ViT-S/16 encoders at every stage, the deployed reader, per-slide predictions and paper data (part 3)

4367928 verified 19 days ago

374 Bytes

y ty sd params label own
0.5890 0.0160 22 {DINOv2-S, natural images (22M)} 0
0.6380 0.0220 26 {ResNet-50, ImageNet (26M)} 0
0.7270 0.0290 1100 {Midnight-12k, ViT-g/14 (1.1B)} 0
0.7490 0.0270 86 {Hibou-B, ViT-B/14 (86M)} 0
0.8050 0.0200 1100 {H-optimus-0, ViT-g/14 (1.1B)} 0
0.8110 0.0140 86 {Phikon, ViT-B/16 (86M)} 0
0.8120 0.0150 22 {Ours, ViT-S/16 (22M)} 1

	y ty sd params label own
	0 0.5890 0.0160 22 {DINOv2-S, natural images (22M)} 0
	1 0.6380 0.0220 26 {ResNet-50, ImageNet (26M)} 0
	2 0.7270 0.0290 1100 {Midnight-12k, ViT-g/14 (1.1B)} 0
	3 0.7490 0.0270 86 {Hibou-B, ViT-B/14 (86M)} 0
	4 0.8050 0.0200 1100 {H-optimus-0, ViT-g/14 (1.1B)} 0
	5 0.8110 0.0140 86 {Phikon, ViT-B/16 (86M)} 0
	6 0.8120 0.0150 22 {Ours, ViT-S/16 (22M)} 1