MolRecBench-Wild: A Real-World Benchmark for Optical Chemical Structure Recognition
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913097514483712 |
|---|---|
| author | Yang, Haote Wang, Hui Zhu, Chen Wang, Jingchao Li, Linye Lai, Hongbin Ao, Huijie Lyu, Yongxuan Wu, Jiang Sun, Jiaxing Chen, Lua Cao, Yuanyuan Zhang, Ruijie Lu, Shengxin Wu, Lijun Wang, Bin He, Conghui |
| author_facet | Yang, Haote Wang, Hui Zhu, Chen Wang, Jingchao Li, Linye Lai, Hongbin Ao, Huijie Lyu, Yongxuan Wu, Jiang Sun, Jiaxing Chen, Lua Cao, Yuanyuan Zhang, Ruijie Lu, Shengxin Wu, Lijun Wang, Bin He, Conghui |
| contents | Optical Chemical Structure Recognition (OCSR) aims to translate molecular diagrams in scientific literature into machine-readable formats, but current systems remain unreliable on real-world images due to substantial visual and chemical complexity. We introduce MOSAIC, a dual-dimensional difficulty framework with 37 fine-grained labels that jointly characterize visual interference and chemical semantic challenges in molecular diagrams. Based on this framework, we construct MolRecBench-Wild, a benchmark of 5,029 structures from 820 recent chemistry papers, covering the full difficulty spectrum observed in real publications. To enable faithful semantic evaluation beyond SMILES and MolFile, we propose CARBON, a representation language capable of expressing valence variations, icon-based groups, and other non-standard chemical semantics. We further adopt a dual-track evaluation protocol supporting both CARBON and SMILES outputs for broad model compatibility. Comprehensive experiments over 18 OCSR-capable models reveal severe performance degradation on MolRecBench-Wild, exposing a large gap between previous patent benchmarks and real-world academic scenarios. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_05832 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MolRecBench-Wild: A Real-World Benchmark for Optical Chemical Structure Recognition Yang, Haote Wang, Hui Zhu, Chen Wang, Jingchao Li, Linye Lai, Hongbin Ao, Huijie Lyu, Yongxuan Wu, Jiang Sun, Jiaxing Chen, Lua Cao, Yuanyuan Zhang, Ruijie Lu, Shengxin Wu, Lijun Wang, Bin He, Conghui Artificial Intelligence Optical Chemical Structure Recognition (OCSR) aims to translate molecular diagrams in scientific literature into machine-readable formats, but current systems remain unreliable on real-world images due to substantial visual and chemical complexity. We introduce MOSAIC, a dual-dimensional difficulty framework with 37 fine-grained labels that jointly characterize visual interference and chemical semantic challenges in molecular diagrams. Based on this framework, we construct MolRecBench-Wild, a benchmark of 5,029 structures from 820 recent chemistry papers, covering the full difficulty spectrum observed in real publications. To enable faithful semantic evaluation beyond SMILES and MolFile, we propose CARBON, a representation language capable of expressing valence variations, icon-based groups, and other non-standard chemical semantics. We further adopt a dual-track evaluation protocol supporting both CARBON and SMILES outputs for broad model compatibility. Comprehensive experiments over 18 OCSR-capable models reveal severe performance degradation on MolRecBench-Wild, exposing a large gap between previous patent benchmarks and real-world academic scenarios. |
| title | MolRecBench-Wild: A Real-World Benchmark for Optical Chemical Structure Recognition |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2605.05832 |