Codesota · Computer Vision · Document Layout Analysis · publaynet-valTasks/Computer Vision/Document Layout Analysis
Document Layout Analysis · benchmark dataset · 2020 · EN

publaynet-val.

Dataset from Papers With Code

Submit a result ↵
§ 01 · Leaderboard

Best published scores.

92 results indexed across 12 metrics. Shaded row marks current SOTA; ties broken by submission date.

Primary
accuracy · higher is better
All metrics
Figure, List, Overall, Table, Text, Title, figure, list, overall, table, text, title
Figure
1 row
#ModelOrgSubmittedPaper / codeFigure
01Hybrid DLA (Shehzadi et al.)OpenDFKI / TU KaiserslauternApr 2024icdar-20240.985
List
1 row
#ModelOrgSubmittedPaper / codeList
01Hybrid DLA (Shehzadi et al.)OpenDFKI / TU KaiserslauternApr 2024icdar-20240.973
Overall
2 rows
#ModelOrgSubmittedPaper / codeOverall
01Hybrid DLA (Shehzadi et al.)OpenDFKI / TU KaiserslauternApr 2024icdar-20240.973
02RoDLAOpenChen, Zhang et al.Mar 2024cvpr-20240.960
Table
1 row
#ModelOrgSubmittedPaper / codeTable
01Hybrid DLA (Shehzadi et al.)OpenDFKI / TU KaiserslauternApr 2024icdar-20240.986
Text
1 row
#ModelOrgSubmittedPaper / codeText
01Hybrid DLA (Shehzadi et al.)OpenDFKI / TU KaiserslauternApr 2024icdar-20240.980
Title
1 row
#ModelOrgSubmittedPaper / codeTitle
01Hybrid DLA (Shehzadi et al.)OpenDFKI / TU KaiserslauternApr 2024icdar-20240.942
figure
14 rows
#ModelOrgSubmittedPaper / codefigure
01DETROpenMeta AI / FAIRJun 2023Bridging the Performance Gap between DETR and R-CNN for …0.975
02DiT-L—Mar 2022DiT: Self-supervised Pre-training for Document Image Tra… · code0.972
03VGT—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.971
04DoPTA—Dec 2024DoPTA: Improving Document Layout Analysis using Patch-Te…0.970
05LayoutLMv3-B—Apr 2022LayoutLMv3: Pre-training for Document AI with Unified Te… · code0.970
06ResNext-101-32×8d—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.968
07TRDLU—Oct 2022papers-with-code0.966
08VSR—May 2021VSR: A Unified Framework for Document Layout Analysis co… · code0.964
09UDoc—Apr 2022Unified Pretraining Framework for Document Understanding0.964
10DeiT-BOpenMetaDec 2020Training data-efficient image transformers & distillatio… · code0.957
11BEiT-B—Jun 2021BEiT: BERT Pre-Training of Image Transformers · code0.957
12Mask R-CNNOpenMeta AI / FAIRAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.949
13Faster R-CNNOpenMicrosoft ResearchAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.937
14GLAM—Aug 2023A Graphical Approach to Document Layout Analysis · code0.206
list
14 rows
#ModelOrgSubmittedPaper / codelist
01TRDLU—Oct 2022papers-with-code0.975
02VGT—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.968
03DETROpenMeta AI / FAIRJun 2023Bridging the Performance Gap between DETR and R-CNN for …0.964
04DiT-L—Mar 2022DiT: Self-supervised Pre-training for Document Image Tra… · code0.960
05DoPTA—Dec 2024DoPTA: Improving Document Layout Analysis using Patch-Te…0.957
06LayoutLMv3-B—Apr 2022LayoutLMv3: Pre-training for Document AI with Unified Te… · code0.955
07VSR—May 2021VSR: A Unified Framework for Document Layout Analysis co… · code0.947
08ResNext-101-32×8d—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.940
09UDoc—Apr 2022Unified Pretraining Framework for Document Understanding0.937
10BEiT-B—Jun 2021BEiT: BERT Pre-Training of Image Transformers · code0.924
11DeiT-BOpenMetaDec 2020Training data-efficient image transformers & distillatio… · code0.921
12Mask R-CNNOpenMeta AI / FAIRAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.886
13Faster R-CNNOpenMicrosoft ResearchAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.883
14GLAM—Aug 2023A Graphical Approach to Document Layout Analysis · code0.862
overall
14 rows
#ModelOrgSubmittedPaper / codeoverall
01VGT—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.962
02TRDLU—Oct 2022papers-with-code0.959
03VSR—May 2021VSR: A Unified Framework for Document Layout Analysis co… · code0.957
04DETROpenMeta AI / FAIRJun 2023Bridging the Performance Gap between DETR and R-CNN for …0.957
05LayoutLMv3-B—Apr 2022LayoutLMv3: Pre-training for Document AI with Unified Te… · code0.951
06DiT-L—Mar 2022DiT: Self-supervised Pre-training for Document Image Tra… · code0.949
07DoPTA—Dec 2024DoPTA: Improving Document Layout Analysis using Patch-Te…0.949
08UDoc—Apr 2022Unified Pretraining Framework for Document Understanding0.939
09ResNext-101-32×8d—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.935
10DeiT-BOpenMetaDec 2020Training data-efficient image transformers & distillatio… · code0.932
11BEiT-B—Jun 2021BEiT: BERT Pre-Training of Image Transformers · code0.931
12Mask R-CNNOpenMeta AI / FAIRAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.910
13Faster R-CNNOpenMicrosoft ResearchAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.902
14GLAM—Aug 2023A Graphical Approach to Document Layout Analysis · code0.722
table
15 rows
#ModelOrgSubmittedPaper / codetable
01DETROpenMeta AI / FAIRJun 2023Bridging the Performance Gap between DETR and R-CNN for …0.981
02VGT—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.981
03LayoutLMv3-B—Apr 2022LayoutLMv3: Pre-training for Document AI with Unified Te… · code0.979
04CDeC-Net—Aug 2020CDeC-Net: Composite Deformable Cascade Network for Table… · code0.978
05DiT-L—Mar 2022DiT: Self-supervised Pre-training for Document Image Tra… · code0.978
06DoPTA—Dec 2024DoPTA: Improving Document Layout Analysis using Patch-Te…0.977
07TRDLU—Oct 2022papers-with-code0.976
08ResNext-101-32×8d—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.976
09VSR—May 2021VSR: A Unified Framework for Document Layout Analysis co… · code0.974
10BEiT-B—Jun 2021BEiT: BERT Pre-Training of Image Transformers · code0.973
11UDoc—Apr 2022Unified Pretraining Framework for Document Understanding0.973
12DeiT-BOpenMetaDec 2020Training data-efficient image transformers & distillatio… · code0.972
13Mask R-CNNOpenMeta AI / FAIRAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.960
14Faster R-CNNOpenMicrosoft ResearchAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.954
15GLAM—Aug 2023A Graphical Approach to Document Layout Analysis · code0.868
text
14 rows
#ModelOrgSubmittedPaper / codetext
01VSR—May 2021VSR: A Unified Framework for Document Layout Analysis co… · code0.967
02TRDLU—Oct 2022papers-with-code0.958
03VGT—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.950
04DETROpenMeta AI / FAIRJun 2023Bridging the Performance Gap between DETR and R-CNN for …0.947
05LayoutLMv3-B—Apr 2022LayoutLMv3: Pre-training for Document AI with Unified Te… · code0.945
06DoPTA—Dec 2024DoPTA: Improving Document Layout Analysis using Patch-Te…0.944
07DiT-L—Mar 2022DiT: Self-supervised Pre-training for Document Image Tra… · code0.944
08UDoc—Apr 2022Unified Pretraining Framework for Document Understanding0.939
09BEiT-B—Jun 2021BEiT: BERT Pre-Training of Image Transformers · code0.934
10DeiT-BOpenMetaDec 2020Training data-efficient image transformers & distillatio… · code0.934
11ResNext-101-32×8d—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.930
12Mask R-CNNOpenMeta AI / FAIRAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.916
13Faster R-CNNOpenMicrosoft ResearchAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.910
14GLAM—Aug 2023A Graphical Approach to Document Layout Analysis · code0.878
title
14 rows
#ModelOrgSubmittedPaper / codetitle
01VGT—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.939
02VSR—May 2021VSR: A Unified Framework for Document Layout Analysis co… · code0.931
03TRDLU—Oct 2022papers-with-code0.921
04DETROpenMeta AI / FAIRJun 2023Bridging the Performance Gap between DETR and R-CNN for …0.918
05LayoutLMv3-B—Apr 2022LayoutLMv3: Pre-training for Document AI with Unified Te… · code0.906
06DoPTA—Dec 2024DoPTA: Improving Document Layout Analysis using Patch-Te…0.895
07DiT-L—Mar 2022DiT: Self-supervised Pre-training for Document Image Tra… · code0.893
08UDoc—Apr 2022Unified Pretraining Framework for Document Understanding0.885
09DeiT-BOpenMetaDec 2020Training data-efficient image transformers & distillatio… · code0.874
10BEiT-B—Jun 2021BEiT: BERT Pre-Training of Image Transformers · code0.866
11ResNext-101-32×8d—Aug 2023Vision Grid Transformer for Document Layout Analysis · code0.862
12Mask R-CNNOpenMeta AI / FAIRAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.840
13Faster R-CNNOpenMicrosoft ResearchAug 2019PubLayNet: largest dataset ever for document layout anal… · code0.826
14GLAM—Aug 2023A Graphical Approach to Document Layout Analysis · code0.800
Fig 2 · Rows sorted by score within each metric. Shaded row marks SOTA. Dates reflect model or paper release where available, otherwise the date Codesota accessed the source.
§ 04 · Literature

12 papers
tied to this benchmark.

Every paper below corresponds to at least one row in the leaderboard above. Click through for the arXiv preprint and, when available, the reference implementation.

§ 06 · Contribute

Have a score that beats
this table?

Submit a checkpoint and a reproduction script. We will run it, publish the score, and — if it takes the top — annotate the step on the progress chart with your name.

Submit a result ↵Read submission guide
What a submission needs
  • 01A public checkpoint or API endpoint
  • 02A reproduction script with frozen commit + seed
  • 03Declared evaluation environment (Python, deps)
  • 04One row per metric declared by this dataset
  • 05A contact so we can follow up on discrepancies