COMO: Closed-Loop Optical Molecule Recognition with Minimum Risk Training

2026-04-26

AI Summary

Optical chemical structure recognition in real-world documents faces significant challenges due to structural diversity, shorthand conventions, and visual noise. Existing approaches are hindered by exposure bias and token-level training objectives, which impede direct optimization of molecule-level evaluation metrics. This work proposes COMO, a novel framework that introduces closed-loop minimum risk training (MRT) to this task for the first time. By leveraging iterative sampling and a self-feedback mechanism, COMO directly optimizes non-differentiable molecular-level objectives such as chemical validity and structural similarity. The method is model-agnostic and seamlessly integrates into end-to-end systems. Evaluated across ten benchmark datasets, COMO substantially outperforms current rule-based and learning-based methods, achieving state-of-the-art performance even with less training data.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Machine Learning: Structured Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Optical chemical structure recognition (OCSR) translates molecular images into machine-readable representations like SMILES strings or molecular graphs, but remains challenging in real-world documents due to inexhaustible variations in chemical structures, shorthand conventions, and visual noise. Most existing deep-learning-based 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 closed-loop framework that mitigates exposure bias by directly optimizing over molecule-level, non-differentiable objectives, by iteratively sampling and evaluating the model's own predictions. 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. Ablation studies further show that MRT is architecture-agnostic, demonstrating its potential for broad application to end-to-end OCSR systems.

Problem

• Optical Chemical Structure Recognition
• Exposure Bias
• Molecule-level Evaluation
• Chemical Validity
• Structural Similarity

Innovation

• Minimum Risk Training
• Closed-loop Framework
• Exposure Bias Mitigation
• Molecule-level Optimization
• Optical Chemical Structure Recognition