J un 2 02 1 ICDAR 2021 Competition on Scientific Literature Parsing

Scientific literature contain important information related to cutting-edge innovations in diverse domains. Advances in natural language processing have been driving the fast development in automated information extraction from scientific literature. However, scientific literature is often available in unstructured PDF format. While PDF is great for preserving basic visual elements, such as characters, lines, shapes, etc., on a canvas for presentation to humans, automatic processing of the PDF format by machines presents many challenges. With over 2.5 trillion PDF documents in existence, these issues are prevalent in many other important application domains as well. A critical challenge for automated information extraction from scientific literature is that documents often contain content that is not in natural language, such as figures and tables. Nevertheless, such content usually illustrates key results, messages, or summarizations of the research. To obtain a comprehensive understanding of scientific literature, the automated system must be able to recognize the layout of the documents and parse the non-natural-language content into a machine readable format. Our ICDAR 2021 Scientific Literature Parsing Competition (ICDAR2021SLP) aims to drive the advances specifically in document understanding. ICDAR2021-SLP leverages the PubLayNet and PubTabNet datasets, which provide hundreds of thousands of training and evaluation examples. In Task A, Document Layout Recognition, submissions with the highest performance combine object detection and specialised solutions for the different categories. In Task B, Table Recognition, top submissions rely on methods to identify table components and post-processing methods to generate the table structure and content. Results from both tasks show an impressive performance and opens the possibility for high performance practical applications.