Salvato in:
Dettagli Bibliografici
Autori principali: Shah, Ayush Kumar, Amador, Bryan Manrique, Dey, Abhisek, Creekmore, Ming, Ocampo, Blake, Denmark, Scott, Zanibbi, Richard
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:https://arxiv.org/abs/2311.12161
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915173827084288
author Shah, Ayush Kumar
Amador, Bryan Manrique
Dey, Abhisek
Creekmore, Ming
Ocampo, Blake
Denmark, Scott
Zanibbi, Richard
author_facet Shah, Ayush Kumar
Amador, Bryan Manrique
Dey, Abhisek
Creekmore, Ming
Ocampo, Blake
Denmark, Scott
Zanibbi, Richard
contents 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. On the synthetic USPTO benchmark, our born-digital parser obtains a recognition rate of 98.4% (1% higher than previous models) and our relatively simple neural parser for raster images obtains a rate of 85% using less training data than existing neural approaches (thousands vs. millions of molecules).
format Preprint
id arxiv_https___arxiv_org_abs_2311_12161
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle ChemScraper: Leveraging PDF Graphics Instructions for Molecular Diagram Parsing
Shah, Ayush Kumar
Amador, Bryan Manrique
Dey, Abhisek
Creekmore, Ming
Ocampo, Blake
Denmark, Scott
Zanibbi, Richard
Computer Vision and Pattern Recognition
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. On the synthetic USPTO benchmark, our born-digital parser obtains a recognition rate of 98.4% (1% higher than previous models) and our relatively simple neural parser for raster images obtains a rate of 85% using less training data than existing neural approaches (thousands vs. millions of molecules).
title ChemScraper: Leveraging PDF Graphics Instructions for Molecular Diagram Parsing
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2311.12161