References & Citations
Computer Science > Computer Vision and Pattern Recognition
Title: ChemScraper: Leveraging PDF Graphics Instructions for Molecular Diagram Parsing
(Submitted on 20 Nov 2023 (v1), last revised 26 Apr 2024 (this version, v3))
Abstract: Most molecular diagram parsers recover chemical structure from raster images (e.g., PNGs). However, many PDFs include commands giving explicit locations and shapes for characters, lines, and polygons. We present a new parser that uses these born-digital PDF primitives as input. The parsing model is fast and accurate, and does not require GPUs, Optical Character Recognition (OCR), or vectorization. We use the parser to annotate raster images and then train a new multi-task neural network for recognizing molecules in raster images. We evaluate our parsers using SMILES and standard benchmarks, along with a novel evaluation protocol comparing molecular graphs directly that supports automatic error compilation and reveals errors missed by SMILES-based evaluation.
Submission history
From: Ayush Kumar Shah [view email][v1] Mon, 20 Nov 2023 20:27:42 GMT (4947kb,D)
[v2] Wed, 22 Nov 2023 03:23:17 GMT (4947kb,D)
[v3] Fri, 26 Apr 2024 16:43:14 GMT (3734kb,D)
Link back to: arXiv, form interface, contact.