Research map / Vision-language and multimodal / Human and embodied modeling
Seeing Symbols, Missing Structure: A Real-World Handwritten Mathematical Expression Recognition Benchmark for Large Models
Sheng Jiang, Lin Zhu, Runrui Li, Mei Wang, Qiannan Zhu, Yaoyao Zhong, Hua Huang
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 |
|---|---|---|---|
| apply full SASR (MSR + SIL) at inference | Qwen3-VL-8B-Instruct · HMER-Bench | ExpRate: 11.63 -> 24.52 | Improved |
| apply SASR | Qwen3-VL-8B-Instruct · HMER-Bench | ExpRate@CDM: 25.84 -> 29.98 | Improved |
| add multi-stage structure-aware reasoning (MSR) (ablation) | Qwen3-VL-8B-Instruct · HMER-Bench | ExpRate: 11.63 -> 19.52 | Improved |
| add schema-anchored in-context learning (SIL) alone (ablation) | Qwen3-VL-8B-Instruct · HMER-Bench | ExpRate: 11.63 -> 14.02 | Improved |
| add SIL on top of MSR (ablation) | Qwen3-VL-8B-Instruct · HMER-Bench | ExpRate: 19.52 -> 24.52 | Improved |
| add Step1 schema inference to MSR (ablation) | Qwen3-VL-8B-Instruct · HMER-Bench | ExpRate: 10.28 -> 19.52 | Improved |
| add Step3 symbol disambiguation to MSR (ablation) | Qwen3-VL-8B-Instruct · HMER-Bench D3 | ExpRate: 5.27 -> 7.39 | Improved |
| add Step3 symbol disambiguation to MSR (ablation) | Qwen3-VL-8B-Instruct · HMER-Bench | ExpRate: 18.84 -> 19.52 | Improved |
| add Simple CoT prompting (ablation) | Qwen3-VL-8B-Instruct · HMER-Bench | ExpRate: 10.99 -> 11.36 | No clear effect |
| apply SASR on Ovis2.5-8B | Ovis2.5-8B · HMER-Bench | ExpRate: 7.72 -> 12.55 | Improved |
| apply SASR on Nemotron-Nano-12B-v2-VL | Nemotron-Nano-12B-v2-VL · HMER-Bench | ExpRate: 6.01 -> 8.91 | Improved |
| apply SASR on Kimi-K2.5 | Kimi-K2.5 · HMER-Bench | ExpRate: 28.06 -> 38.90 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Uni-MuMER — Uni-MuMER: Unified Multi-Task Fine-Tuning of Vision-Language Model for Handwritten Mathematical Expression Recognition (NeurIPS 2025) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Uni-MuMER: Unified Multi-Task Fine-Tuning of Vision-Language Model for Handwritten Mathematical Expression Recognition (NeurIPS 2025 · similarity 0.92)
- SSAN: A Symbol Spatial-Aware Network for Handwritten Mathematical Expression Recognition (AAAI 2025 · similarity 0.91)
- TAMER: Tree-Aware Transformer for Handwritten Mathematical Expression Recognition (AAAI 2025 · similarity 0.89)
- Generating Handwritten Mathematical Expressions From Symbol Graphs: An End-to-End Pipeline (CVPR 2024 · similarity 0.88)
- OpenWebMath: An Open Dataset of High-Quality Mathematical Web Text (ICLR 2024 · similarity 0.83)
- Primitive Vision: Improving Diagram Understanding in MLLMs (ICML 2025 · 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.
