List your company
ResearchFormerly WizWand

Semantic Segmentation on ADE20K 70 (val)

59.4mIoUMask2Former
55.96856.85957.7558.641Mar 24, 2025Jun 9, 2025Aug 25, 2025Nov 11, 2025Jan 27, 2026Apr 14, 2026Jul 1, 2026
Updated 4mo ago

Evaluation results

MethodLinks
Mask2FormerBackbone=ViT-L [23], Pre-training=DINOv2 + DA, Input size=896x8962025.0359.4——
Mask2FormerBackbone=ViT-L [17], Pre-training=DINOv2 + DA [62], Input size=896²2026.0759.4——
Mask2FormerBackbone=ViT-Adapter-L [13], Pre-training=DINOv2, Params=351M, Input size=512x512, GFLOPs=910, FPS=21, re-implementation=true2025.0358.9——
Mask2FormerBackbone=ViT-Adapter-L [10], Pre-training=DINOv2, Params=351M, Input size=512²2026.0758.991021
LUMABackbone=ViT-L [17], Pre-training=DINOv2, Params=320M, Input size=512²2026.0758.7719154
EoMTBackbone=ViT-L [23], Pre-training=DINOv2, Params=319M, Input size=512x512, GFLOPs=721, FPS=922025.0358.4——
EoMT‡Backbone=ViT-L [17], Pre-training=DINOv2, Params=319M, Input size=512²2026.0758.4721150
OneFormerBackbone=DiNAT-L [30], Pre-training=IN21K, Params=223M, Input size=896x896, GFLOPs=678, FPS=192025.0358.1——
OneFormer†Backbone=DiNAT-L [22], Pre-training=IN21K, Params=223M, Input size=896²2026.0758.167819
OneFormerBackbone=ConvNext-XL [46], Pre-training=IN21K, Params=373M, Input size=640x640, GFLOPs=607, FPS=212025.0357.4——
OneFormer†Backbone=ConvNext-XL [38], Pre-training=IN21K, Params=373M, Input size=640²2026.0757.460721
MaskDINOBackbone=Swin-L [45], Pre-training=IN21K, Params=223M, Input size=640x6402025.0356.6——
MaskDINO†Backbone=Swin-L [37], Pre-training=IN21K, Params=223M, Input size=640²2026.0756.6——
Mask2FormerBackbone=Swin-L [45], Pre-training=IN21K, Params=216M, Input size=640x640, FPS=332025.0356.1——
Mask2Former†Backbone=Swin-L [37], Pre-training=IN21K, Params=216M, Input size=640²2026.0756.1—33