Research map / Detection and segmentation / CNN features
PIX-TAB: Efficient PIXel-Precise TABle Structure Recognition Approach with Speculative Decoding and Region-Based Image Segmentation
Viktor Zaytsev, Olena Vynokurova, Pavlo Tytarchuk, Dmytro Kozii, Vitalii Pohribnyi, Olga Radyvonenko, Artem Shcherbina
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add speculative decoding (ablation) | PIX-TAB · FinTabNet | FPS: 3.80 -> 7.96 | Improved |
| add speculative decoding (ablation) | PIX-TAB · PubTabNet | FPS: 3.36 -> 8.54 | Improved |
| add speculative decoding (ablation) | PIX-TAB · FinTabNet | TEDS-struct: 98.72 -> 98.72 | No clear effect |
| add region-based image segmentation (RBIS) (ablation) | PIX-TAB · SynthTabNet MarketingStyle | TEDS100: 35.08 -> 45.61 | Improved |
| add RBIS (ablation) | PIX-TAB · SynthTabNet MarketingStyle | TEDS-struct100: 56.14 -> 57.59 | Improved |
| add RBIS (ablation) | PIX-TAB · FinTabNet | TEDS-struct: 98.72 -> 98.65 | Hurt |
| add RBIS with speculative decoding (ablation) | PIX-TAB · FinTabNet | FPS: 7.96 -> 7.23 | Hurt |
| add RBIS without speculative decoding (ablation) | PIX-TAB · PubTabNet | FPS: 3.36 -> 2.82 | Hurt |
| add proposed synthetic Wikipedia-based data (ablation) | PIX-TAB · FinTabNet | TEDS100: 77.30 -> 77.51 | Improved |
| add proposed synthetic data (ablation) | PIX-TAB · PubTabNet | TEDS100: 70.49 -> 70.60 | Improved |
| add proposed synthetic data instead of SynthTabNet (ablation) | PIX-TAB · FinTabNet | TEDS-struct100: 89.79 -> 89.83 | No clear effect |
| use optimized mobile model | PIX-TAB · PubTabNet on mobile | avg time (s): 19.9 -> 6.6 | Improved |
Closest papers by meaning
- Grab What You Need: Rethinking Complex Table Structure Recognition with Flexible Components Deliberation (AAAI 2024 · similarity 0.86)
- Table2LaTeX-RL: High-Fidelity LaTeX Code Generation from Table Images via Reinforced Multimodal Language Models (NeurIPS 2025 · similarity 0.82)
- TDeLTA: A Light-Weight and Robust Table Detection Method Based on Learning Text Arrangement (AAAI 2024 · similarity 0.82)
- Decoupling Skeleton and Flesh: Efficient Multimodal Table Reasoning with Disentangled Alignment and Structure-aware Guidance (ICML 2026 · similarity 0.82)
- Multimodal Tabular Reasoning with Privileged Structured Information (NeurIPS 2025 · similarity 0.82)
- TRivia: Self-supervised Fine-tuning of Vision-Language Models for Table Recognition (CVPR 2026 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
