Papers should be formatted according to the instructions and style files provided by Springer. The LaTeX template for LNCS can be downloaded here. It is also available on Overleaf.

Papers are limited to 15 pages (not including references). Papers that do not meet these requirements will be desk rejected. The accepted papers for ICDAR 2025 will be presented in Oral or Poster presentation forms. All papers should be submitted electronically via the online submission form here (https://cmt3.research.microsoft.com/ICDAR2025).

Be careful, complete, and accurate when submitting papers. No changes in paper authorship will be permitted after a paper has been submitted.

No-show papers (papers not presented at the conference in person by one of the authors) will not be eligible for a best paper award, and the author(s) of the paper will not receive a conference participation certificate. If an author is unable to attend in person due to an unsuccessful visa application or other insurmountable difficulties, they should immediately contact the conference organizers and the Program Chairs who will determine whether an online presentation can be made instead. In any case, whether the presentation is in person or online, authors must obey the same rules regarding the conference registration process.

Topics of Interest:

Document image processing Structured document generation
Physical and logical layout analysis Multimedia document analysis
Text and symbol recognition Mobile text recognition
Handwriting recognition Pen-based document analysis
Document analysis systems Scene text detection and recognition
Document classification Recognition of tables and formulas
Indexing and retrieval of documents Historical document analysis
Document synthesis Signature verification
Extracting document semantics Document summarization and translation
NLP for document understanding Document forensics and provenance
Office automation Medical document analysis
Graphics recognition Document analysis for social good
Human document interaction Document analysis for literature search
Document Representation Modeling Gold-standard benchmarks and datasets
Document VQA Foundation models (Large Multimodal
Private and secure Models) for document understanding document understanding

Publication: Springer Lecture Notes in Computer Science

The conference proceedings will be published as part of the Springer Lecture Notes in Computer Science (LNCS) series. Accepted papers will be freely available through SpringerLink from the conference website for one year after publication, and will later be freely available through SpringerLink four years after publication.

Contact: 

Xu-Cheng Yin  <xuchengyin@ustb.edu.cn>

Dimosthenis Karatzas <dimos@cvc.uab.es>

Daniel Lopresti <lopresti@cse.lehigh.edu>