Research map / Detection and segmentation / Open-set and few-shot detection
Table Structure Recognition using Top-Down and Bottom-Up Cues
Sachin Raja, Ajoy Mondal, C V Jawahar
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 alignment loss to Mask R-CNN+TD+BU with DGCNN+P2+LSTM (ablation) | TabStruct-Net · ICDAR-2013 (S-A, table image only) | structure recognition F1: 0.888 -> 0.906 | Improved |
| add alignment loss (ablation) | TabStruct-Net · ICDAR-2013 (S-A, table image only) | cell detection F1: 0.907 -> 0.924 | Improved |
| add top-down and bottom-up FPN pathways to Mask R-CNN (ablation) | TabStruct-Net · ICDAR-2013 (S-A, table image only) | structure recognition F1: 0.882 -> 0.888 | Improved |
| add LSTM visual features to DGCNN+P2 (ablation) | TabStruct-Net · ICDAR-2013 (S-A, table image only) | structure recognition F1: 0.870 -> 0.882 | Improved |
| add P2 features to DGCNN (ablation) | TabStruct-Net · ICDAR-2013 (S-A, table image only) | structure recognition F1: 0.865 -> 0.870 | Improved |
| add LSTM to DGCNN+P2 (ablation) | TabStruct-Net · ICDAR-2013 (S-B, with cell boxes) | structure recognition F1: 0.978 -> 0.981 | Improved |
| add P2 features to DGCNN (ablation) | TabStruct-Net · ICDAR-2013 (S-B, with cell boxes) | structure recognition F1: 0.977 -> 0.978 | Improved |
| raise IoU threshold from 0.6 to 0.7 (ablation) | TabStruct-Net · ICDAR-2013 (S-A, table image only) | structure recognition F1: 0.906 -> 0.791 | Hurt |
| give ground-truth cell boxes as input (S-B instead of S-A) | TabStruct-Net · ICDAR-2013 | structure recognition F1: 0.906 -> 0.981 | Improved |
| train on ICDAR-2019 instead of SciTSR | TabStruct-Net · ICDAR-2019 test | structure recognition F1: 0.583 -> 0.785 | Improved |
| replace SPLERGE with TabStruct-Net | TabStruct-Net · ICDAR-2013 (S-A, table image only) | structure recognition F1: 0.879 -> 0.906 | Improved |
Closest papers by meaning
- TGRNet: A Table Graph Reconstruction Network for Table Structure Recognition (ICCV 2021 · similarity 0.90)
- Grab What You Need: Rethinking Complex Table Structure Recognition with Flexible Components Deliberation (AAAI 2024 · similarity 0.88)
- TDeLTA: A Light-Weight and Robust Table Detection Method Based on Learning Text Arrangement (AAAI 2024 · similarity 0.87)
- TRivia: Self-supervised Fine-tuning of Vision-Language Models for Table Recognition (CVPR 2026 · similarity 0.84)
- Graph-based Document Structure Analysis (ICLR 2025 · similarity 0.83)
- M2Doc: A Multi-Modal Fusion Approach for Document Layout Analysis (AAAI 2024 · similarity 0.83)
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.
