bioRxiv 10k with assets

<p>This dataset is a superset of the <a href="https://doi.org/10.5281/zenodo.3873702">bioRxiv 10k dataset</a>. It additionally includes the assets (usually images) that are linked by the XML files.</p> <p>The assets were retrieved from the <a href="https://www.biorxiv.org/tdm">bioRxiv&#39;s tdm bucket</a>. Assets that are not linked by the XML were omitted. Examples of those would be some large videos.</p> <p>This dataset serves a similar purpose as the <a href="https://doi.org/10.5281/zenodo.3873702">bioRxiv 10k dataset</a>, but for use-cases that require the assets. e.g. training and evaluation of figure image extraction.</p> <p>This dataset mirrors the exact same documents and structure as the <a href="https://doi.org/10.5281/zenodo.3873702">bioRxiv 10k dataset</a>. But rather than just containing the PDF and XML files, it also contains the linked assets (often images, but not necessarily).</p> <p>The dataset was created as part of <a href="https://elifesciences.org/">eLife</a>&#39;s <a href="https://github.com/elifesciences/sciencebeam">ScienceBeam</a> project.</p>