Tables Decoded: DELTA for Structure, TARQA for Understanding

Table understanding is a core task in document intelligence, encompassing two key subtasks: table reconstruction and table visual question answering (TabVQA). While recent approaches predominantly rely on vision-language models (VLMs) operating on table images, we propose a more scalable and effective alternative based on structured textual representations. These representations are easier to process, align naturally with large language models (LLMs), and remove the need for language-specific visual encoders — making them ideal for multilingual documents. We present DELTA, a decoupled table reconstruction framework that separates structure recognition from OCR to extract both layout and content accurately. DELTA outputs tables in Optimized Table Structure Language (OTSL), a compact and unified format that encodes cell arrangements and textual content. DELTA achieves high-fidelity table-to-text conversion, outperforming prior methods on structure metrics with superior TEDS-Structure scores across FinTabNet, PubTabNet, and PubTables.
Built on DELTA, we introduce TARQA (Table structure-Aware Representation for Question Answering) — an LLM fine-tuned on OTSL-formatted tables for accurate and structure-aware TabVQA. TARQA outperforms baselines fine-tuned on HTML representations by 14.2%, and improves answer accuracy on WTQ by 6.8% and FinTabNetQA by 9.2%.

├── Ablation/        # Ablation experiments and analysis
├── DELTA/           # DELTA: Table reconstruction module
├── TARQA/           # TARQA: Structure-aware table QA module
├── TORQUE/          # TORQUE dataset sample images and scripts
├── TabVQA/          # Tabular question answering datasets and configs

• Text-based table representation → more scalable than VLM-based approaches.
• Multilingual compatibility → no dependence on visual encoders.
• Superior structure reconstruction → highest TEDS-Structure scores on FinTabNet, PubTabNet, and PubTables.
• Structure-aware QA → TARQA improves TabVQA accuracy by +14.2% over HTML baselines.

git clone https://github.com/JahanviRajputTIH/Tables-Understanding.git
cd Tables-Understanding
pip install -r requirements.txt

Located under Ablation/, covering:

• OCR–layout decoupling effect
• Impact of OTSL vs HTML representations
• Multilingual fine-tuning analysis
• Structure fidelity vs reasoning accuracy trade-off

If you use this paper or the accompanying code/data in your research, please cite it as:

@inproceedings{tablesdecoded2025,
  title={Tables Decoded: DELTA for Structure, TARQA for Understanding},
  author={Jahanvi Rajput and Dhruv Kudale and Saikiran Kasturi and Utkarsh Verma},
  booktitle={Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)},
  year={2025}
}

This project is licensed under the MIT License. See LICENSE for details.

1. Jahanvi Rajput
2. Dhruv Kudale
3. Saikiran Kasturi
4. Utkarsh Verma