COMO：用最小风险训练实现闭环光学分子识别

View PDF HTML (experimental)

Abstract:Optical chemical structure recognition (OCSR) translates molecular images into machine-readable representations such as SMILES strings or molecular graphs, but remains challenging in real-world documents due to wide variations in chemical structures, shorthand conventions, and scanning artifacts. Most existing deep learning approaches rely on teacher forcing with token-level Maximum Likelihood Estimation (MLE). This training paradigm suffers from exposure bias, as models are trained under ground-truth prefixes but must condition on their own previous predictions during inference. Moreover, token-level MLE objectives hinder the optimization towards molecular-level evaluation criteria such as chemical validity and structural similarity. Here we introduce Minimum Risk Training (MRT) to OCSR and propose COMO (Closed-loop Optical Molecule recOgnition), a framework that mitigates exposure bias by directly optimizing over molecule-level, non-differentiable objectives. Experiments on ten benchmarks, including synthetic and real-world chemical diagrams from patent and scientific literature, demonstrate that COMO substantially outperforms existing rule-based and learning-based methods with less training data, consistently achieving superior performance across both reproduced baselines and previously reported results. Ablation studies further show that MRT is architecture-agnostic, demonstrating its potential for broad application to end-to-end OCSR systems.

Subjects:	Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
Cite as:	arXiv:2604.23546 [cs.CV]
	(or arXiv:2604.23546v2 [cs.CV] for this version)
	https://doi.org/10.48550/arXiv.2604.23546 arXiv-issued DOI via DataCite

Submission history

From: Zhuoqi Lyu [view email]
[v1] Sun, 26 Apr 2026 05:48:21 UTC (801 KB)
[v2] Tue, 10 Nov 2026 08:56:30 UTC (2,309 KB)