VisionDocs

Overview

In this age of progressive digitalization, the ability to analyze documents in an automated way is gaining increasing importance in our everyday lives. The impact of Document Analysis is becoming more and more prominent both in industrial as well as cultural settings, which leads to the necessity to develop AI systems that are capable of analyzing documents characterized by highly heterogeneous features, going from the differences in the languages used, due to the different geographical and temporal origins of the documents, to the way the documents appear from a visual perspective, with different writing styles and layouts being adopted across different domains.

In recent years, significant advancements have been made in the field of Document Analysis, particularly within computer vision, but much of the progress has focused on a limited subset of document types and tasks. This leaves many areas, especially those involving low-resource languages, non-standard layouts, and historical documents, under-explored. The existing research landscape has yet to fully address issues such as the ability to generalize across diverse document formats or to operate effectively in low-data scenarios where training examples are scarce. Moreover, integrating multimodal inputs (e.g., combining text with visual, structural, or semantic information) to enhance document understanding is an ongoing challenge that needs further exploration.

Document Analysis is a multidisciplinary space that brings together a wide range of data sources, application domains, and involved disciplines. This workshop aims to promote cross-disciplinary interactions, facilitating knowledge exchange and collaborative efforts. By fostering collaboration between experts from various fields, we hope to drive innovation and advance the development of Document Analysis solutions.

Call for Paper

Research papers are solicited in, but not limited to, the following topic areas:

Keynote Speaker 

Submission

We invite researchers to submit their original and unpublished work related to the workshop's theme. Authors can submit either full papers (max 8 pages + reference) or short papers (max 4 pages + reference), following the ICCV 2025 formatting guidelines. Accepted full papers will published under the ICCV 2025 workshop proceedings.

All submissions should be compiled for single-blind review, adopt the standard main conference ICCV 2025 template.

Accepted papers will be presented during the workshop as oral presentations or posters.

Submission site: https://openreview.net/group?id=thecvf.com/ICCV/2025/Workshop/VisionDocs  

Important Dates

• Regular Papers:

  • Paper submissions: 15 June 29 June, 2025 23:59 UTC-0

  • Author Notification: 11 July, 2025 23:59 UTC-0

  • Camera-ready: 30 July, 2025 23:59 UTC-0

• Short Papers and Demos:

  • Paper submissions: 17 August, 2025 23:59 UTC-0

  • Author Notification: 05 September, 2025 23:59 UTC-0

  • Short paper Camera-ready: 14 September, 2025 23:59 UTC-0

• Workshop date: October 20th afternoon, 2025