DECIMER

The official project page of DECIMER: Deep Learning for Chemical Image Recognition

Description

The project aims to develop methods to employ deep learning to recognize and interpret chemical structure from images in the printed and online literature with the aim of re-discovering scientific facts about natural products and their meta-data. The method, of course, will be generically applicable to most organic chemistry publications.

The project is co-supervised by,

• Prof. Christoph Steinbeck, Friedrich Schiller University, Jena.
• Prof. Achim Zielesny, Westphalian University of Applied Sciences, Recklinghausen.

Project Repository

• DECIMER - OCSR Review
  • An initial work done by us to review and benchmark the open-source OCSR(Optical Chemical Structure Recognition) systems available
• DECIMER - Image to Smiles(Pre release - Preliminary report available)
  • The repository contains the network and the related scripts for auto-encoder based Chemical Image Recognition
• DECIMER - Python (Developmental stage)
  • This repository hosts the python scripts written throughout the project.
• DECIMER - Java (Developmental stage)
  • This repository hosts the JAVA scripts written throughout the project.
• CUDA installation instructions
  • This repository hosts the step by step guide to install nVidia Drivers, CUDA drivers and Tensorflow 2.0

License:

• This project is licensed under the MIT License - see the LICENSE file for details

Author:

• Kohulan Rajan
  • github page
  • Profile

Project Website

• DECIMER

Research Group

• Website