# POTATR: A Lightweight Image-to-Graph Model for Page-Level Table Extraction

arXiv: https://arxiv.org/abs/2606.09788 · HTML: https://mlsanity.com/paper/2606.09788/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 31 references that are in the map.
- First posted on arXiv 8 Jun 2026.

## Builds on (most cited first, 24 of 24)

- [End-to-End Object Detection with Transformers](https://mlsanity.com/paper/2005.12872/) (2020)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Flamingo: a Visual Language Model for Few-Shot Learning](https://mlsanity.com/paper/2204.14198/) (2022)
- [Deformable DETR: Deformable Transformers for End-to-End Object Detection](https://mlsanity.com/paper/2010.04159/) (2020)
- [Image-based table recognition: data, model, and evaluation](https://mlsanity.com/paper/1911.10683/) (2019)
- [Power Hungry Processing: Watts Driving the Cost of AI Deployment?](https://mlsanity.com/paper/2311.16863/) (2023)
- [DeepSeek-OCR: Contexts Optical Compression](https://mlsanity.com/paper/2510.18234/) (2025)
- [Global Table Extractor (GTE): A Framework for Joint Table Identification and Cell Structure Recognition Using Visual Context](https://mlsanity.com/paper/2005.00589/) (2020)
- [PaddleOCR 3.0 Technical Report](https://mlsanity.com/paper/2507.05595/) (2025)
- [PubTables-1M: Towards comprehensive table extraction from unstructured documents](https://mlsanity.com/paper/2110.00061/) (2021)
- [olmOCR: Unlocking Trillions of Tokens in PDFs with Vision Language Models](https://mlsanity.com/paper/2502.18443/) (2025)
- [CascadeTabNet: An approach for end to end table detection and structure recognition from image-based documents](https://mlsanity.com/paper/2004.12629/) (2020)
- [TableFormer: Table Structure Understanding with Transformers](https://mlsanity.com/paper/2203.01017/) (2022)
- [DeepSeek-OCR 2: Visual Causal Flow](https://mlsanity.com/paper/2601.20552/) (2026)
- [TableNet: Deep Learning model for end-to-end Table detection and Tabular data extraction from Scanned Document Images](https://mlsanity.com/paper/2001.01469/) (2020)
- [Relationformer: A Unified Framework for Image-to-Graph Generation](https://mlsanity.com/paper/2203.10202/) (2022)
- [SmolDocling: An ultra-compact vision-language model for end-to-end multi-modal document conversion](https://mlsanity.com/paper/2503.11576/) (2025)
- [EGTR: Extracting Graph from Transformer for Scene Graph Generation](https://mlsanity.com/paper/2404.02072/) (2024)
- [dots.ocr: Multilingual Document Layout Parsing in a Single Vision-Language Model](https://mlsanity.com/paper/2512.02498/) (2025)
- [GriTS: Grid table similarity metric for table structure recognition](https://mlsanity.com/paper/2203.12555/) (2022)
- [Granite Vision: a lightweight, open-source multimodal model for enterprise Intelligence](https://mlsanity.com/paper/2502.09927/) (2025)
- [Aligning benchmark datasets for table structure recognition](https://mlsanity.com/paper/2303.00716/) (2023)
- [TRACE: Table Reconstruction Aligned to Corner and Edges](https://mlsanity.com/paper/2305.00630/) (2023)
- [PubTables-v2: A new large-scale dataset for full-page and multi-page table extraction](https://mlsanity.com/paper/2512.10888/) (2025)

## Abstract (arXiv)

Large-scale document processing requires contextually aware table extraction (TE) that is both accurate and efficient. Yet current approaches require billions of parameters, hundreds of autoregressive steps, or costly API inference. Motivated by this, we introduce the Page-Object Table Transformer (POTATR), a lightweight 29M parameter image-to-graph model that extends the Table Transformer (TATR) for contextualized page-level TE. POTATR outperforms all models tested on the PubTables-v2 Single Pages benchmark -- including frontier MLLMs -- achieving textrmGriTS_textrmCon of 0.964 while running over 130× faster at roughly 300× lower cost. Further, POTATR's output is spatially grounded: every recognized element has a bounding box, enabling visual verification and geometric text assignment. As a result, POTATR performs unified page-level TE while composing with other models, enabling extension to scanned documents via external OCR and to full-document TE via techniques like cross-page merging. Code and models will be released.