Important Dates
About the Challenge
Challenge Overview
The Forensic Handwritten Document Analysis (FHDA) Challenge invites participants to tackle the binary classification task of determining whether a given pair of handwritten documents were authored by the same individual.
This competition introduces a unique cross-modal comparison between traditional handwritten documents (scanned) and documents written directly on digital devices such as tablets or graphic tablets.
Challenge Significance
The FHDA Challenge holds paramount importance in forensic science, writer identification, and document authentication.
Precise authorship verification is essential for:
- Detecting document forgery
- Ensuring the integrity of legal documents
- Strengthening forensic investigations
- Historical manuscript studies
Challenge Objective
Develop innovative solutions or architectures to achieve higher accuracy in authorship verification across different writing modalities. The dataset encompasses diverse handwriting styles, writing instruments, and environmental conditions, making it representative of real-world forensic challenges.
Evaluation Criteria
Models will be evaluated based on a composite score calculated by multiplying the number of correct classifications by the assigned probability for each comparison. This weighted accuracy approach rewards both correctness and appropriate confidence levels in predictions.
Dataset Information
A novel dataset will be exclusively released for this challenge, consisting of:
- Handwritten documents written on paper and later scanned
- Documents written directly on digital devices
- Paired documents for comparative analysis
The dataset will encompass diverse handwriting styles, writing instruments, and environmental conditions, making it representative of real-world forensic handwriting analysis challenges.
Dataset Access and Details: The dataset download links, detailed specifications, and any updates will be made available on the official challenge GitHub repository: https://github.com/mfs-iplab/fhda-challenge. Participants are encouraged to monitor the repository for the latest information.
Scoring System
For each pair of documents, participants must:
- Determine whether the documents were written by the same author (binary classification)
- Provide a confidence percentage for each comparison
The final score will be calculated by multiplying the number of correct classifications by the assigned probability for each comparison. The team with the highest overall score wins the challenge.
How to Participate
Registration Guidelines
Teams must register no later than May 16, 2025 by completing the registration form on this website. Please provide:
- Team name
- Team members (first name, last name, affiliation, email contact)
Note: It will not be allowed to unify teams after the beginning of the challenge, even if they belong to the same university.
Register Your TeamSubmission Requirements
Participants must submit their results no later than June 20, 2025. The submission must include:
Technical Documentation
- Detailed description of the proposed architecture
- Dataset management approach
- Brief description of the methodology
- Preliminary results obtained
Results Format
- Written in English
- Complete details of the approach
- Parameters used in the model
- Comparison with baseline methods
- Follow the provided templates
Final rankings and ground truth will be released on June 25, 2025.
The winning team will be invited to submit a paper describing their approach by July 20, 2025.
Download Templates (Link TBD)Registered Teams
Below are the teams that have registered for the FHDA Challenge 2025. The list will be updated as new teams register.
Institution: Zayed University, Abu Dhabi
Team Email: richard.ikuesan@zu.ac.ae
Team Referent: Richard Ikuesan
Team Members:
- Richard Ikuesan, Nigeria, Zayed University
- Hessa Almazrouei, UAE, Zayed University
- Thani Al-Riyami Alremeithi, Oman, Zayed University
- Rahaf Alnuaimi, UAE, Zayed University
- Khalifa, UAE, Zayed University
Institution: Zayed University, Abu Dhabi
Team Email: richard.ikuesan@zu.ac.ae
Team Referent: Richard Ikuesan
Team Members:
- Richard Ikuesan, Nigeria, Zayed University
- Noura Alzaabi, UAE, Zayed University
- Maryam Almarzooqi, UAE, Zayed University
- Shaikha Alzaabi, UAE, Zayed University
- Noor Aldahmani, UAE, Zayed University
Publication Opportunities
- The winning team will have the opportunity to submit a paper that, if accepted after a review process, will be considered in the conference proceedings
- Authors of the best works will be invited to present at the IEEE MetroXRAINE 2025 conference
- The description of the challenge will be submitted to a top journal in the field
- Paper submission deadline is July 20, 2025
Main Contact
Mirko Casu
Ph.D. Student
University of Catania, Italy
IEEE Student Member
For all inquiries regarding the challenge, please contact:
Email: mirko.casu@phd.unict.it
Challenge Organizers
Frequently Asked Questions
What is the main goal of this challenge?
The main goal is to develop innovative solutions for authorship verification in handwritten documents, particularly focusing on cross-modal comparison between traditional pen-and-paper documents and digitally written samples.
Who can participate in the challenge?
The challenge is open to researchers, developers, and students from academic institutions and industry with interest in document analysis, pattern recognition, and machine learning.
How will submissions be evaluated?
Submissions will be evaluated using a scoring system that multiplies the number of correct authorship classifications by the confidence percentage assigned to each comparison. This rewards both accuracy and appropriate confidence levels in your predictions.
What should my submission include?
Your submission must include your architecture details, dataset management approach, a brief methodology description, and preliminary results. All submissions must be in English and follow the provided templates.
Where can I find the dataset and related resources?
All dataset download links, detailed specifications, and updates will be available on the official challenge GitHub repository: https://github.com/mfs-iplab/fhda-challenge.
