Codesota · Computer Vision · Table Recognition · pubtabnetTasks/Computer Vision/Table Recognition
Table Recognition · benchmark dataset · 2020 · EN

pubtabnet.

Dataset from Papers With Code

Submit a result ↵
§ 01 · Leaderboard

Best published scores.

30 results indexed across 2 metrics. Shaded row marks current SOTA; ties broken by submission date.

Primary
accuracy · higher is better
All metrics
teds-all-samples, teds-struct
teds-all-samples
16 rows
#ModelOrgSubmittedPaper / codeteds-all-samples
01SEMv3OpenIFLYTEK / USTC (Zhang et al.)May 2024SEMv3: A Fast and Robust Approach to Table Separation Li…97.30
02MuTabNet—Apr 2024Multi-Cell Decoder and Mutual Learning for Table Structu… · code96.87
03TABLETOpenAnonymous (arXiv 2025)Jun 2025TABLET: Table Structure Recognition using Encoder-only T…96.79
04TableMaster—May 2021PingAn-VCGroup's Solution for ICDAR 2021 Competition on … · code96.76
05Multi-Task Learning Model—Mar 2023An End-to-End Multi-Task Learning Model for Image-based … · code96.67
06TFLOPOpenUpstage AIJan 2025TFLOP: Table Structure Recognition Framework with Layout…96.66
07ConvStem—Nov 2023High-Performance Transformers for Table Structure Recogn… · code96.53
08UniTable LargeOpenGeorgia Tech (Peng et al.)Mar 2024UniTable: Towards a Unified Framework for Table Recognit… · code96.50
09SLANet—Oct 2022PP-StructureV2: A Stronger Document Analysis System · code96.30
10TRUST—Aug 2022TRUST: An Accurate and End-to-End Table structure Recogn…96.20
11GridFormerOpenAnonymous (arXiv 2023)Sep 2023GridFormer: Towards Accurate Table Structure Recognition…95.84
12NCGM—Nov 2021Neural Collaborative Graph Machines for Table Structure …95.40
13LGPMA—May 2021LGPMA: Complicated Table Structure Recognition with Loca… · code94.60
14SEM—Jul 2021Split, embed and merge: An accurate table structure reco…93.70
15TabStruct-Net—Oct 2020Table Structure Recognition using Top-Down and Bottom-Up… · code90.10
16EDD—Nov 2019Image-based table recognition: data, model, and evaluati… · code88.30
teds-struct
14 rows
#ModelOrgSubmittedPaper / codeteds-struct
01TFLOPOpenUpstage AIJan 2025TFLOP: Table Structure Recognition Framework with Layout…98.38
02UniTable LargeOpenGeorgia Tech (Peng et al.)Mar 2024UniTable: Towards a Unified Framework for Table Recognit… · code97.89
03Multi-Task Learning Model—Mar 2023An End-to-End Multi-Task Learning Model for Image-based … · code97.88
04TABLETOpenAnonymous (arXiv 2025)Jun 2025TABLET: Table Structure Recognition using Encoder-only T…97.67
05UniTabNetOpenAnonymous / ACL communitySep 2024UniTabNet: Bridging Vision and Language Models for Enhan…97.50
06TSRFormer—Aug 2022TSRFormer: Table Structure Recognition with Transformers97.50
07SEMv3OpenIFLYTEK / USTC (Zhang et al.)May 2024SEMv3: A Fast and Robust Approach to Table Separation Li…97.50
08TRUST—Aug 2022TRUST: An Accurate and End-to-End Table structure Recogn…97.10
09SLANet—Oct 2022PP-StructureV2: A Stronger Document Analysis System · code97.01
10GridFormerOpenAnonymous (arXiv 2023)Sep 2023GridFormer: Towards Accurate Table Structure Recognition…97
11RTSR—Mar 2022Robust Table Detection and Structure Recognition from He…97
12LGPMA—May 2021LGPMA: Complicated Table Structure Recognition with Loca… · code96.70
13SPRINTOpenIIT Bombay LEAP-OCRMar 2025SPRINT: Script-agnostic Structure Recognition in Tables · code95.71
14TabStruct-Net—Oct 2020Table Structure Recognition using Top-Down and Bottom-Up… · code90.10
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

20 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