Call For Papers! The Inaugural ICDAR Workshop on Document Image and Language (DIL 2021)


Back to list

图片 1.png

Baidu, together with the Institute of Automation of the Chinese Academy of Sciences (CASIA), the German Research Center for Artificial Intelligence (DFKI), and the State University of New York at Buffalo, will collectively hold DIL 2021 – namely, the Workshop on Document Image and Language – at the International Conference on Document Analysis and Recognition (ICDAR) 2021 for the first time. We sincerely invite scholars and experts of relevant disciplines to participate in this workshop.


ICDAR is revered as one of the most influential and important international conferences in the field of document analysis and recognition. The event is organized by the International Association for Pattern Recognition (IAPR) and held every two years. Since its establishment, it has focused on the research and application of cutting-edge technology in the field of document recognition. The ICDAR vigorously promotes academic exchanges among many international universities and enterprises, also facilitating the commercialization of relevant research achievements and technologies in education, medical treatments, office management, social networking, information retrieval and other fields that will ultimately benefit mankind through science and technology.


Document Image Analysis and Recognition (DIAR) has always been the primary focus of the ICDAR conference. In the past, researchers of this community usually used computer vision technology alone to address the problems of text detection and recognition, layout analysis, and even graphic structure recognition.


Today, multi-modality in artificial intelligence has become a promising trend and natural language processing (NLP) technology has gained rapid development in recent years. In order to understand the abundant multimodal information of documents, the various techniques from NLP, such as Named Entity Recognition and Linking, Visual Questions Answering, and Text Classification, can be combined with highly established traditional methods, such as OCR, Layout Analysis, and Logical Labeling. 


To this end, Baidu is holding the inaugural DIL 2021 workshop. Following and promoting the tradition of ICDAR workshops, we are striving to accelerate the application of text modality in document image understanding by providing a forum for exchange and fostering a union between these two fields. By bringing together associate researchers from document image understanding, the NLP community, and those who are at the forefront of engaging in research that infuses both disciplines, workshop participants can collectively share their professional knowledge and latest findings. This can generate more insights and culminate in a more thorough understanding of document multimodal information.


The website of the first workshop of DIL2021 is here:


We sincerely invite scholars and experts to submit papers on the topics below:



OCR with Semantic Model

l  Post-OCR text correction

l  Semantic text recognition

l  Semantic text spotting

NLP Technologies on Document

l  Document entity recognition

l  Document entity relationship

l  Document visual question answering (DocVQA)

l  Text classification

l  Message and table understanding

l  Document sentiment analysis

l  Text similarity measures

l  Fact or relationship extraction

Multi-Modality Document Image Analysis and Recognition

l  Pretrained model for document analysis

l  Document information extraction

l  Document layout understanding

l  Graphic and formula recognition in document

l  Medical document image recognition

l  Document classification

l  Document summarization and translation


Link for paper submission:


Paper Submission Dates:

l  Workshop paper submission deadline: May 17

l  Workshop paper notificationJune 21

l  Workshop camera-ready paper submissions: July 5



l  Andreas Dengel, DFKI & University of Kaiserslautern, Germany

l Cheng-Lin Liu, Institute of Automation of Chinese Academy of Sciences (CASIA), China

l  David Doermann, University of Buffalo, USA

l  Errui Ding, Baidu Inc., China

l  Hua Wu, Baidu Inc., China

l  Jingtuo Liu, Baidu Inc., China