Research map / Probabilistic and scientific modeling / Molecules and proteins
MolParser: End-to-end Visual Recognition of Molecule Structures in the Wild
Xi Fang, Jiankun Wang, Xiaochen Cai, Shangqian Chen, Shuwen Yang, Haoyi Tao, Nan Wang, Lin Yao, Linfeng Zhang, Guolin Ke
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 |
|---|---|---|---|
| pretrain on MolParser-7M instead of MolGrapher-300k (ablation) | MolParser · WildMol-10K | accuracy: 22.4 -> 51.9 | Improved |
| add fine-tuning on in-the-wild data after pretraining (ablation) | MolParser · WildMol-10K | accuracy: 51.9 -> 76.9 | Improved |
| separate pretraining then fine-tuning instead of one mixed stage (ablation) | MolParser · WildMol-10K | accuracy: 75.9 -> 76.9 | Improved |
| add data augmentation (ablation) | MolParser · WildMol-10K | accuracy: 40.1 -> 69.5 | Improved |
| add curriculum learning strategy (ablation) | MolParser · WildMol-10K | accuracy: 69.5 -> 76.9 | Improved |
| use InternViT-300M encoder (2200M params) instead of Swin-Base (ablation) | MolParser · WildMol-10K | accuracy: 76.9 -> 72.9 | Hurt |
| scale from Swin-Tiny 224 to Swin-Base 384 | MolParser · WildMol-10K | accuracy: 73.1 -> 76.9 | Improved |
| scale from Swin-Tiny to Swin-Base | MolParser · WildMol-10K-M | accuracy: 15.3 -> 38.1 | Improved |
| use MolParser pretrained Swin-T instead of ImageNet pretrained | Swin-T · MoleculeNet 5 tasks avg | ROC-AUC: 68.9 -> 73.7 | Improved |
| use MolParser-Tiny instead of Base for speed | MolParser · WildMol throughput | images per second: 39.8 -> 131.6 | Improved |
Closest papers by meaning
- MarkushGrapher: Joint Visual and Textual Recognition of Markush Structures (CVPR 2025 · similarity 0.92)
- Atom-Level Optical Chemical Structure Recognition with Limited Supervision (CVPR 2024 · similarity 0.91)
- MolVision: Molecular Property Prediction with Vision Language Models (NeurIPS 2025 · similarity 0.88)
- RFL: Simplifying Chemical Structure Recognition with Ring-Free Language (AAAI 2025 · similarity 0.88)
- TinyChemVL: Advancing Chemical Vision-Language Models via Efficient Visual Token Reduction and Complex Reaction Tasks (AAAI 2026 · similarity 0.86)
- ChemVLM: Exploring the Power of Multimodal Large Language Models in Chemistry Area (AAAI 2025 · similarity 0.86)
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.
