Research map / Label-efficient and robust learning / Recognition and long-tail
TDv2: A Novel Tree-Structured Decoder for Offline Mathematical Expression Recognition
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use TDv2's tree-structured decoder instead of DLA | HMER · CROHME 2014 | ExpRate: 51.88 -> 53.62 | Helped |
| Use TDv2's tree-structured decoder instead of DLA | HMER · CROHME 2016 | ExpRate: 51.53 -> 55.18 | Helped |
| Use TDv2's tree-structured decoder instead of DenseWAP-TD | HMER · CROHME 2019 | ExpRate: 51.4 -> 58.72 | Helped |
| Add thinking labels (ablation) | TDv2 · CROHME 2014 | ExpRate: 49.46 -> 53.65; 4.19% | Helped |
| Add thinking labels (ablation) | TDv2 · CROHME 2016 | ExpRate: 47.32 -> 50.87; 3.55% | Helped |
| Add thinking labels (ablation) | TDv2 · CROHME 2019 | ExpRate: 51.90 -> 55.91; 4.01% | Helped |
| Add pixel-level auxiliary classification loss (ablation) | TDv2 with thinking labels · CROHME 2014 | ExpRate: 53.65 -> 53.62; -0.03% | Hurt |
| Add pixel-level auxiliary classification loss (ablation) | TDv2 with thinking labels · CROHME 2016 | ExpRate: 50.87 -> 54.16; 3.29% | Helped |
| Add pixel-level auxiliary classification loss (ablation) | TDv2 with thinking labels · CROHME 2019 | ExpRate: 55.91 -> 58.72; 2.81% | Helped |
Closest papers by meaning
- TAMER: Tree-Aware Transformer for Handwritten Mathematical Expression Recognition (AAAI 2025 · similarity 0.92)
- Uni-MuMER: Unified Multi-Task Fine-Tuning of Vision-Language Model for Handwritten Mathematical Expression Recognition (NeurIPS 2025 · similarity 0.89)
- SSAN: A Symbol Spatial-Aware Network for Handwritten Mathematical Expression Recognition (AAAI 2025 · similarity 0.87)
- Generating Handwritten Mathematical Expressions From Symbol Graphs: An End-to-End Pipeline (CVPR 2024 · similarity 0.86)
- A Neural-Guided Dynamic Symbolic Network for Exploring Mathematical Expressions from Data (ICML 2024 · similarity 0.81)
- MathCoder2: Better Math Reasoning from Continued Pretraining on Model-translated Mathematical Code (ICLR 2025 · similarity 0.80)
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.