List your company
ResearchFormerly WizWand

Instance Segmentation on COCO

54.8APmaskBEIT-3
42.21645.48348.7552.017Mar 19, 2021Oct 28, 2021Jun 9, 2022Jan 19, 2023Aug 31, 2023Apr 11, 2024Nov 21, 2024
Updated 4mo ago

Evaluation results

MethodLinks
BEIT-32022.0854.8——
Mask DINO2022.0854.7——
Uni-Perceiver-v2 LARGEparameters=446M, task-specific fine-tuning=false2022.1153.6——
HIPIEBackbone=ViT-H2023.0751.9——
HIPIEBackbone=ViT-H2023.0751.9——
Uni-Perceiver-v2 BASEparameters=308M, task-specific fine-tuning=false2022.1150.6——
MViTv2-L (SoftNMS, MS)Backbone=MViTv2-L, Pre-training=IN-21K, Training schedule=50 epochs, large-scale jittering, Inference strategy=SoftNMS and multiscale testing, Params=270 M2021.1250.5——
MViTv2-H (IN-21K, LSJ)Backbone=MViTv2-H, Pre-training=IN-21K, Training schedule=50 epochs, large-scale jittering, FLOPs=3084 G, Params=718 M2021.1248.5——
MViTv2-L (IN-21K, LSJ)Backbone=MViTv2-L, Pre-training=IN-21K, Training schedule=50 epochs, large-scale jittering, FLOPs=1519 G, Params=270 M2021.1248.3——
CAE*Backbone=ViT-L, #Epochs=1600, Pre-training Supervision=Self-supervised, Framework=Mask R-CNN, Schedule=1x2022.0247.672.251.9
CAEBackbone=ViT-L, #Epochs=1600, Pre-training Supervision=Self-supervised, Framework=Mask R-CNN, Schedule=1x2022.0247.67251.9
MViTv2-B (IN-21K)Backbone=MViTv2-B, Pre-training=IN-21K, FLOPs=814 G, Params=103 M2021.1247.4——
TEC_MAEBackbone=ViT-B, Framework=Cascade MaskRCNN, Source Implementation=Li et al., 2022b2022.1047.2——
MViTv2-LBackbone=MViTv2-L, FLOPs=1519 G, Params=270 M2021.1247.1——
MAEBackbone=ViT-L, #Epochs=1600, Pre-training Supervision=Self-supervised, Framework=Mask R-CNN, Schedule=1x2022.0247.171.551
VIC-MAEpre-train data=IN1K+K710+MiT, Backbone=ViT-B Mask R-CNN2023.0346.9——
MViTv2-BBackbone=MViTv2-B, FLOPs=814 G, Params=103 M2021.1246.8——
SEEMBackbone=DaViT-B2023.0746.8——
SEEMBackbone=DaViT-B2023.0746.8——
MAEBackbone=ViT-B, Framework=Cascade MaskRCNN, Source Implementation=Li et al., 2022b2022.1046.7——
C-MAEpre-train data=IN1K, Backbone=ViT-B Mask R-CNN2023.0346.5——
VIC-MAEpre-train data=IN1K+K400, Backbone=ViT-B Mask R-CNN2023.0346.5——
SiameseIM2022.0646.2——
MViTv2-SBackbone=MViTv2-S, FLOPs=748 G, Params=87 M2021.1246——
ODISEBackbone=ViT-H+SD2023.0746——
ODISEBackbone=ViT-H+SD2023.0746——
RDNet-BBackbone=RDNet-B, Param=144M, FLOPs=971G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.034669.150
MAE2022.0645.9——
HIPIEBackbone=RN502023.0745.9——
HIPIEBackbone=RN502023.0745.9——
X-DecoderBackbone=DaViT-B2023.0745.8——
X-DecoderBackbone=DaViT-B2023.0745.8——
ConvNeXt-BBackbone=ConvNeXt-B, Param=146M, FLOPs=964G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.0345.668.949.5
TEC_iBOTBackbone=ViT-B, Framework=Cascade MaskRCNN, Source Implementation=Zhou et al., 2022a2022.1045.4——
RDNet-SBackbone=RDNet-S, Param=108M, FLOPs=832G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.0345.468.549.3
C-JEPAPretrain Epochs=600, Backbone=ViT-B/162024.1045.3——
NAT-BBackbone=NAT-B, Param=147M, FLOPs=931G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.0345.268.649
Full-precisionNetwork=Swin-S + Cascade Mask R-CNN, #Bits=32/32, Size=427.82024.1145——
Full-precisionNetwork=Swin-B + Cascade Mask R-CNN, #Bits=32/32, Size=579.92024.1145——
RepQ-ViT + QwTNetwork=Swin-B + Cascade Mask R-CNN, #Bits=6/6, Size=126.12024.1145——
MViTv2-TBackbone=MViTv2-T, FLOPs=701 G, Params=76 M2021.1245——
Swin-BBackbone=Swin-B, FLOPs=982 G, Params=145 M2021.1245——
Swin-SBackbone=Swin-S, Param=107M, FLOPs=838G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.034568.248.8
ConvNeXt-SBackbone=ConvNeXt-S, Param=108M, FLOPs=827G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.034568.449.1
Swin-BBackbone=Swin-B, Param=145M, FLOPs=982G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.034568.148.9
NAT-SBackbone=NAT-S, Param=108M, FLOPs=809G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.0344.968.148.6
MAEpre-train data=IN1K, Backbone=ViT-B Mask R-CNN2023.0344.9——
MAEPretrain Epochs=1600, Backbone=ViT-B/162024.1044.9——
RepQ-ViT + QwTNetwork=Swin-S + Cascade Mask R-CNN, #Bits=6/6, Size=91.32024.1144.8——
RepQ-ViTNetwork=Swin-B + Cascade Mask R-CNN, #Bits=6/6, Size=112.12024.1144.8——
Swin-SBackbone=Swin-S, FLOPs=838 G, Params=107 M2021.1244.7——
RepQ-ViTNetwork=Swin-S + Cascade Mask R-CNN, #Bits=6/6, Size=83.42024.1144.6——
HPMPretrain Epoch=800, Pre-trained Data=IN1K, Resolution=224x2242024.0744.6——
RDNet-TBackbone=RDNet-T, Param=81M, FLOPs=757G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.0344.667.948.3
NAT-TBackbone=NAT-T, Param=85M, FLOPs=737G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.0344.567.647.9
I-JEPAPretrain Epochs=600, Backbone=ViT-B/162024.1044.5——
ColorMAE-GPretrain Epoch=1600, Pre-trained Data=IN1K, Resolution=224x2242024.0744.467.848
BEiTPretrain Epochs=800, Backbone=ViT-B/162024.1044.4——
mc-BEiTIntermediate Fine-tuning=True (ImageNet-1K), Task head=Cascaded Mask R-CNN, Training schedule=1x2022.0344.3——
iBOTArch.=ViT-B/162021.1144.2——
iBOTIntermediate Fine-tuning=False, Task head=Cascaded Mask R-CNN, Training schedule=1x2022.0344.2——
iBOTBackbone=ViT-B, Framework=Cascade MaskRCNN, Source Implementation=Zhou et al., 2022a2022.1044.2——
CAEBackbone=ViT-B, #Epochs=1600, Pre-training Supervision=Self-supervised, Framework=Mask R-CNN, Schedule=1x2022.0244.268.347.9
iBOTPretrain Epochs=1600, Backbone=ViT-B/162024.1044.2——
CAE*Backbone=ViT-B, #Epochs=1600, Pre-training Supervision=Self-supervised, Framework=Mask R-CNN, Schedule=1x2022.024467.947.6
CAE-GReaTDetector=Mask R-CNN, Schedule=1x2024.074467.947.3
CAE*Backbone=ViT-B, #Epochs=800, Pre-training Supervision=Self-supervised, Framework=Mask R-CNN, Schedule=1x2022.0243.967.847.4
CSwin-BBackbone=CSwin-B, Param=97M, FLOPs=526G, Framework=Mask-RCNN, Schedule=1x, Image size=(1280, 800)2024.0343.967.847.3
BEITIntermediate Fine-tuning=True (ImageNet-1K), Task head=Cascaded Mask R-CNN, Training schedule=1x2022.0343.8——
MAEPretrain Epoch=1600, Pre-trained Data=IN1K, Resolution=224x2242024.0743.866.647.5
RepQ-ViT + QwTNetwork=Swin-B + Cascade Mask R-CNN, #Bits=4/4, Size=90.12024.1143.7——
Swin-TBackbone=Swin-T, FLOPs=745 G, Params=86 M2021.1243.7——
ColorMAE-GPretrain Epoch=800, Pre-trained Data=IN1K, Resolution=224x2242024.0743.767.147.1
Swin-TBackbone=Swin-T, Param=86M, FLOPs=745G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.0343.766.647.3
ConvNeXt-TBackbone=ConvNeXt-T, Param=86M, FLOPs=741G, Schedule=3x, Image size=1280x800, Framework=Cascade Mask-RCNN2024.0343.766.547.3
BEiTArch.=ViT-B/162021.1143.5——
MixedAEPretrain Epoch=800, Pre-trained Data=IN1K, Resolution=224x2242024.0743.566.247.4
RepQ-ViT + QwTNetwork=Swin-S + Cascade Mask R-CNN, #Bits=4/4, Size=64.82024.1143.4——
DINOArch.=ViT-B/162021.1143.4——
DINOIntermediate Fine-tuning=False, Task head=Cascaded Mask R-CNN, Training schedule=1x2022.0343.4——
MAEPretrain Epoch=800, Pre-trained Data=IN1K, Resolution=224x2242024.0743.466.646.9
RDNet-BBackbone=RDNet-B, Param=107M, FLOPs=493G, Framework=Mask-RCNN, Schedule=1x, Image size=(1280, 800)2024.0343.467.546.6
DINOPretrain Epochs=1600, Backbone=ViT-B/162024.1043.4——
Full-precisionNetwork=Swin-S + Mask R-CNN, #Bits=32/32, Size=276.52024.1143.3——
FocalNet-B (SRF)Backbone=FocalNet-B (SRF), Param=109M, FLOPs=496G, Framework=Mask-RCNN, Schedule=1x, Image size=(1280, 800)2024.0343.367.546.5
SEERData=IG-1B, Params=693 M2021.0343.2——
SupervisedArch.=ViT-B/162021.1143.2——
CSwin-SBackbone=CSwin-S, Param=54M, FLOPs=342G, Framework=Mask-RCNN, Schedule=1x, Image size=(1280, 800)2024.0343.267.146.2
RepQ-ViT + QwTNetwork=Swin-S + Mask R-CNN, #Bits=6/6, Size=61.22024.1143.1——
RepQ-ViTNetwork=Swin-S + Cascade Mask R-CNN, #Bits=4/4, Size=56.92024.1143.1——
RepQ-ViTNetwork=Swin-B + Cascade Mask R-CNN, #Bits=4/4, Size=76.12024.1143.1——
mc-BEiTIntermediate Fine-tuning=False, Task head=Cascaded Mask R-CNN, Training schedule=1x2022.0343.1——
DetCon_BData=IN-1M, Params=250 M2021.0343——
RDNet-SBackbone=RDNet-S, Param=70M, FLOPs=354G, Framework=Mask-RCNN, Schedule=1x, Image size=(1280, 800)2024.034366.946.3
RepQ-ViTNetwork=Swin-S + Mask R-CNN, #Bits=6/6, Size=53.32024.1142.9——
SupervisedIntermediate Fine-tuning=False, Task head=Cascaded Mask R-CNN, Training schedule=1x2022.0342.9——
BEITIntermediate Fine-tuning=False, Task head=Cascaded Mask R-CNN, Training schedule=1x2022.0342.8——
MoCo v3Intermediate Fine-tuning=False, Task head=Cascaded Mask R-CNN, Training schedule=1x2022.0342.7——
MoCo-v32022.0642.7——
iBOTBackbone=ViT-B, #Epochs=1600+, Pre-training Supervision=Self-supervised, Framework=Mask R-CNN, Schedule=1x2022.0242.766.546