List your company

Document Layout Analysis on PubLayNet (val)

0.978mAPCDeC-Net++
0.724240.790120.8560.92188Aug 25, 2020Jan 25, 2021Jun 27, 2021Nov 27, 2021Apr 29, 2022Sep 29, 2022Mar 1, 2023

Evaluation results

MethodLinks
CDeC-Net++Training Dataset=IIIT-AR-13K, Training #Image=9K, Fine-tuning Dataset=PubLayNet, Fine-tuning #Image=340K, IoU=0.5-0.92020.080.9780.9750.9930.984——————————————————————
CDeC-NetTraining Dataset=PubLayNet, Training #Image=340K, IoU=0.5-0.92020.080.9670.970.9880.978——————————————————————
M-RCNNTraining Dataset=PubLayNet, Training #Image=340K, IoU=0.5-0.92020.080.96—————————————————————————
VSRdataset_split=val2021.050.957———0.9670.9310.9470.9740.964—————————————————
StrucTexTv2Model size=Large2023.030.955—————————————————————————
F-RCNNTraining Dataset=PubLayNet, Training #Image=340K, IoU=0.5-0.92020.080.954—————————————————————————
StrucTexTv2Model size=Small2023.030.954—————————————————————————
LayoutLMv3_BASEFramework=Cascade R-CNN, Backbone=Transformer2022.040.951————————0.9450.9060.9550.9790.97————————————
Layout LMv3Backbone size=Base2023.030.951—————————————————————————
DiTBackbone size=Large2023.030.949—————————————————————————
DiT_BASEFramework=Cascade R-CNN, Backbone=Transformer2022.040.945————————0.9440.8890.9480.9760.969————————————
PP-StructureV2Backbone=PP-PicoDet, Speed (ms)=41.22022.100.94—————————————————————————
UDocFramework=Faster R-CNN, Backbone=ResNet-502022.040.939————————0.9390.8850.9370.9730.964————————————
UDoc2023.030.939—————————————————————————
DiT_BASEFramework=Mask R-CNN, Backbone=Transformer2022.040.935————————0.9340.8710.9290.9730.967————————————
BEITBackbone size=Base2023.030.931—————————————————————————
Mask RCNNdataset_split=val2021.050.91———0.9160.840.8860.960.949—————————————————
PubLayNetFramework=Mask R-CNN, Backbone=ResNet-1012022.040.91————————0.9160.840.8860.960.949————————————
Faster RCNNdataset_split=val2021.050.902———0.910.8260.8830.9540.937—————————————————
layoutparserBackbone=Detectron2, Speed (ms)=2900.02022.100.8898—————————————————————————
CDeC-Net++Training Dataset=IIIT-AR-13K, Training #Image=9K, IoU=0.5-0.92020.080.7340.7670.7850.776——————————————————————
[46]Architecture=Mask R-CNN2022.03———0.910.9160.840.8860.960.949—————————————————
BEiT-BBackbone=BEiT-B, Architecture=Mask R-CNN2022.03———0.9310.9340.8660.9240.9730.957—————————————————
BEiT-B2024.12————————————————————0.9340.8660.9240.9730.9570.931
DeiT-BBackbone=DeiT-B, Architecture=Mask R-CNN2022.03———0.9320.9340.8740.9210.9720.957—————————————————
DeiT-B2024.12————————————————————0.9340.8740.9210.9720.9570.932
DiT-BBackbone=DiT-B, Architecture=Mask R-CNN2022.03———0.9350.9340.8710.9290.9730.967—————————————————
DiT-BParameters=87M2024.12————————————————————0.9440.8890.9480.9760.9690.945
DiT-B (C)Backbone=DiT-B, Architecture=Cascade Mask R-CNN2022.03———0.9450.9440.8890.9480.9760.969—————————————————
DiT-LBackbone=DiT-L, Architecture=Mask R-CNN2022.03———0.9410.9370.8790.9450.9740.968—————————————————
DiT-LParameters=304M2024.12————————————————————0.9440.8930.960.9780.9720.949
DiT-L (C)Backbone=DiT-L, Architecture=Cascade Mask R-CNN2022.03———0.9490.9440.8930.960.9780.972—————————————————
Donut-EncoderParameters=72M2024.12————————————————————0.9390.8750.9520.9760.9690.942
DOPTAParameters=85M2024.12————————————————————0.9440.8950.9570.9770.970.949
GLAM2023.08——————————————0.8780.80.8620.8680.2060.722——————
LayoutLM v32023.08——————————————0.9450.9060.9550.9790.970.951——————
LayoutLMv3-BaseParameters=133M, OCR as input=true, Longer finetuning schedule=true2024.12————————————————————0.9450.9060.9550.9790.970.951
M2DocOCR as input=true2024.12————————————————————0.9430.8870.9520.9730.9670.945
MAE-BBackbone=MAE-B, Architecture=Mask R-CNN2022.03———0.930.9330.8650.9180.9730.959—————————————————
MAE-B2024.12————————————————————0.9330.8650.9180.9730.9590.93
MBC2023.08——————————————0.8880.5280.8810.9770.840.823——————
ResNextBackbone=ResNext-101-32x8d, Architecture=Mask R-CNN2022.03———0.920.9160.8450.9180.9710.952—————————————————
ResNeXt2024.12————————————————————0.9160.8450.9180.9710.9520.92
ResNext (C)Backbone=ResNext-101-32x8d, Architecture=Cascade Mask R-CNN2022.03———0.9350.930.8620.940.9760.968—————————————————
StructTextv2-LargeParameters=238M, OCR as input=true, Longer finetuning schedule=true2024.12—————————————————————————0.955
UDoc2024.12————————————————————0.9390.8850.9370.9730.9640.939
VGTParameters=174M, OCR as input=true2024.12————————————————————0.9480.9280.9530.9770.9670.955
VSR2023.08——————————————0.9670.9310.9470.9740.9640.957——————
YOLO v5x62023.08——————————————0.9170.7910.8690.9650.9470.898——————