<?xml version="1.0" encoding="utf-8"?>
<export-example>
  <doc>
    <id>4541</id>
    <completedYear>2024</completedYear>
    <publishedYear>2024</publishedYear>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue>16: 78</issue>
    <volume>2024</volume>
    <type>article</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>2024-07-15</completedDate>
    <publishedDate>2024-07-05</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Advancements in hand-drawn chemical structure recognition through an enhanced DECIMER architecture</title>
    <abstract language="eng">Accurate recognition of hand-drawn chemical structures is crucial for digitising hand-written chemical information in traditional laboratory notebooks or facilitating stylus-based structure entry on tablets or smartphones. However, the inherent variability in hand-drawn structures poses challenges for existing Optical Chemical Structure Recognition (OCSR) software. To address this, we present an enhanced Deep lEarning for Chemical ImagE Recognition (DECIMER) architecture that leverages a combination of Convolutional Neural Networks (CNNs) and Transformers to improve the recognition of hand-drawn chemical structures. The model incorporates an EfficientNetV2 CNN encoder that extracts features from hand-drawn images, followed by a Transformer decoder that converts the extracted features into Simplified Molecular Input Line Entry System (SMILES) strings. Our models were trained using synthetic hand-drawn images generated by RanDepict, a tool for depicting chemical structures with different style elements. A benchmark was performed using a real-world dataset of hand-drawn chemical structures to evaluate the model's performance. The results indicate that our improved DECIMER architecture exhibits a significantly enhanced recognition accuracy compared to other approaches.</abstract>
    <parentTitle language="eng">Journal of Cheminformatics</parentTitle>
    <identifier type="doi">10.1186/s13321-024-00872-7</identifier>
    <licence>© Alle Rechte vorbehalten</licence>
    <author>Kohulan Rajan</author>
    <author>Henning Otto Brinkhaus</author>
    <author>Christoph Steinbeck</author>
    <author>Achim Zielesny</author>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>DECIMER</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>Hand-drawn chemical structures</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>OCSR, Optical Chemical Structure Recognition</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>Transformer</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>Deep Learning</value>
    </subject>
    <collection role="institutes" number="">Institut für biologische und chemische Informatik</collection>
    <thesisPublisher>Westfälische Hochschule Gelsenkirchen Bocholt Recklinghausen</thesisPublisher>
  </doc>
  <doc>
    <id>4542</id>
    <completedYear>2024</completedYear>
    <publishedYear>2024</publishedYear>
    <thesisYearAccepted/>
    <language>eng</language>
    <pageFirst/>
    <pageLast/>
    <pageNumber/>
    <edition/>
    <issue/>
    <volume/>
    <type>preprint</type>
    <publisherName/>
    <publisherPlace/>
    <creatingCorporation/>
    <contributingCorporation/>
    <belongsToBibliography>0</belongsToBibliography>
    <completedDate>2024-07-15</completedDate>
    <publishedDate>--</publishedDate>
    <thesisDateAccepted>--</thesisDateAccepted>
    <title language="eng">Advancements in Hand-Drawn Chemical Structure Recognition through an Enhanced DECIMER Architecture</title>
    <abstract language="eng">Accurate recognition of hand-drawn chemical structures is crucial for digitising hand-written chemical information found in traditional laboratory notebooks or for facilitating stylus-based structure entry on tablets or smartphones. However, the inherent variability in hand-drawn structures poses challenges for existing Optical Chemical Structure Recognition (OCSR) software. To address this, we present an enhanced Deep lEarning for Chemical ImagE Recognition (DECIMER) architecture that leverages a combination of Convolutional Neural Networks (CNNs) and Transformers to improve the recognition of hand-drawn chemical structures. The model incorporates an EfficientNetV2 CNN encoder that extracts features from hand-drawn images, followed by a Transformer decoder that converts the extracted features into Simplified Molecular Input Line Entry System (SMILES) strings. Our models were trained using synthetic hand-drawn images generated by RanDepict, a tool for depicting chemical structures with different style elements. To evaluate the model's performance, a benchmark was performed using a real-world dataset of hand-drawn chemical structures. The results indicate that our improved DECIMER architecture exhibits a significantly enhanced recognition accuracy compared to other approaches.</abstract>
    <identifier type="doi">10.26434/chemrxiv-2024-7ch9f</identifier>
    <enrichment key="opus.doi.autoCreate">false</enrichment>
    <enrichment key="opus.urn.autoCreate">true</enrichment>
    <licence>© Alle Rechte vorbehalten</licence>
    <author>Kohulan Rajan</author>
    <author>Henning Otto Brinkhaus</author>
    <author>Achim Zielesny</author>
    <author>Christoph Steinbeck</author>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>OCSR, Optical Chemical Structure Recognition</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>DECIMER</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>Deep Learning</value>
    </subject>
    <subject>
      <language>eng</language>
      <type>uncontrolled</type>
      <value>Transformer</value>
    </subject>
    <collection role="institutes" number="">Institut für biologische und chemische Informatik</collection>
    <thesisPublisher>Westfälische Hochschule Gelsenkirchen Bocholt Recklinghausen</thesisPublisher>
  </doc>
</export-example>
