Multi-class Semantic Segmentation on ACDC (Dice and IoU per Class) benchmark leaderboard · SOTA2 Research

Multi-class Semantic Segmentation on ACDC (Dice and IoU per Class)

91.23Dice RVU-Net Transformer

Evaluation Results

Method	Links
U-Net Transformer2026.06		91.23	89.34	96.64	92.4	83.87	80.73	93.5	86.03	91.85	92.97	92.28	96.25
MS-DKC-UTrans-ABR2026.06		90.86	89.28	96.83	92.32	83.26	80.64	93.85	85.91	91.63	93.04	92.21	96.25
U-Net2026.06		90.72	89.04	96.65	92.14	83.02	80.25	93.52	85.6	91.76	92.56	92.02	96.11
DKC-HDNeXt-Lite2026.06		85.12	82.32	93.83	87.09	74.1	69.95	88.37	77.47	86.77	87.52	86.9	93.26
Shallow U-Net2026.06		77.66	81.61	90.04	83.1	63.48	68.93	81.89	71.43	82.89	83.43	82.85	88.43