Summer Challenge on Writer Verification, under NCVPRIPG'23

Challenge Updates

  • Registrations are now closed
  • Semi-final dataset released
  • Semi-finals over
  • Top-6 qualifying teams announced
  • Write-up submission closed
  • Event schedule updated
  • Challenge Convened

Challenge gallery


Welcome to the Summer Challenge on Writer Verification, hosted as part of the National Conference on Computer Vision, Pattern Recognition, Image Processing and Graphics (NCVPRIPG'23). Given a pair of handwritten Hindi Text images, the task is to automatically determine whether they are written by the same writer or different writers.

Same Writer

Writer Verification task originated as a signature verification problem to flag potentials frauds in banking-sector. This is a challenging task as the individual handwriting can significantly vary from person to person and thus the underlying model needs to learn the varations in the handwriting style. The problem becomes even more complex in the offline setting, where the dynamic information about the writing process is not available (like writing in an electronic equipment).

Evaluation metrics

Following the literature on verification tasks, we shall use the following performance measures:

  • F1
  • AUC
The F1 score is a measure of a model's accuracy that takes into account both precision and recall.
F1 score = 2 * (precision * recall) (precision + recall)

The Area Under the Curve (AUC) is a metric that measures the performance of a binary classification model by calculating the area under the receiver operating characteristic (ROC) curve. The AUC ranges from 0 to 1, where a score of 1 indicates a perfect classifier, while a score of 0.5 indicates a random classifier. Note that the Leaderboard will be decided based on the AUC score only.

Input and output specifications

Training set contains 1352 folders of images. All the images in one folder are written by the same person. Validation set contains a set of images from 92 different writers. The val.csv file contains name of image pairs and the corresponding labels. A label of 1 indicates that the images are written by the same writer and a label of 0 indicates that the images are written by different writers.
For futher details, please refer to the Github repository.

Test set contains images from 360 writers. In test.csv you are given name of image pairs. For output, you need to predict the label for given pair of images and submit the csv file in the format by editing the test.csv file. Note that test set will be released at a later date than train and validation set.

Code to get started

A github repository is provided to get started with the challenge. The repository contains a sample code to read the dataset and evaluate the performance and other necessary instructions.


Dataset description here

The dataset for consists of handwritten text images in Hindi. The problem is to identify whether writer of a given pair of images are same or not. The dataset is divided into three folders: train, test, and val.

  • The train folder contains 1352 subfolders, where each subfolder represents a different writer. The images in each subfolder are samples of handwritten text written by that particular writer.
  • The val folder contains 730 handwritten text images from 92 writers. Additionally, you will be provided with a val.csv file, which contains the labels for the validation data. Each row in val.csv contains the names of an image pairs followed by a label (1 or 0), indicating whether the two images were written by the same writer or different writers.
  • The test folder contains 2926 handwritten text images from 360 writers. You will also be provided with a test.csv file, which does not contain any labels. Your task is to predict the writer for each image in the test dataset.
Before using the dataset, please read the disclaimer provided here.

Training Dataset

Google Drive link will be provided in the mail

Validation Dataset

Google Drive link will be provided in the mail

Testing Dataset

Google Drive link will be provided in the mail


Test leaderboard

Team AUC
MaSha 0.976
InkSq 0.922
Neural Survivor 0.915
Forza Code 0.912
Alpha 0.878
Word Detective 0.820
Baseline 0.650

Events schedule

When: July 21, 2023; 1730-1930 IST

Time slot (in IST) Activity Speaker
1730-1735 Welcome and Introduction Organizing team
1735-1745 Task Dataset and Challenge Overview Dr. Anand Mishra
1745-1815 Invited Talk: "Catch me if you can; Writer verification in offline signatures" Dr. Sounak Dey, AI Research Engineer, Helsing AI
1815-1845 Invited talk : "Elastic matching algorithms for online signature verification" Dr. Suresh Sundaram, IIT Guwahati
1845-1850 Winner Announcement Dr. Anand Mishra
1850-1920 Team presentations ~
1920-1925 Closing remarks Dr. Anand Mishra

Events timeline

Event Deadline
Challenge announced April 14, 2023
Release of Training and validation Data April 14, 2023
Registration Closes April 30, 2023
Release of Test Data (for semifinal) June 29, 2023
Last date for submitting the results (for semifinal) June 30, 2023
Last date for submitting the trained model June 30, 2023
Last date for submitting the inference model June 30, 2023
Leaderboard updates (for semifinal) July 1, 2023
Last date to submit a writeup about the submitted solution July 10, 2023
Release of test data (for finals) July 16, 2023
Last date for submitting the results (for final) July 16, 2023
Announcement of final winners July 21-23, 2023 (during conference)
Closing of the competetion July 23, 2023


Registrations are now closed.

Challenge Rules

  • This challenge is open to all (students and professionals)
  • Participants can either register as a solo particiapnt or can form a team.
  • Rules for team formation:
    • A team can have a maximum of 4 participants
    • Team members can be from same or different organizations/affiliations
    • A participant can only be a part of a single team
    • Only one member from the team has to register for the challenge
    • One team can only have one registration. Multiple registrations can lead to disqualification.
    • There is no limitations on the number of teams from the same organizations/affiliations (However, one participant can only be part of an unique team)
  • Data download will be permitted only after the team has completed the registration
  • Attending conference (NCVPRIPG'23) will be highly encouraged. Only attending team will get the certificates and awards.
  • The evaluation will be done in two stages: Semifinal (Online) and Final (in-person). Only top-performing teams will be invited to participate in the final evaluation that will be held during the conference. In the case of ties, the organizing committee may rank teams based on the method’s novelty and readability of codes. The organizing committee’s decision in this regard will be final.

Awards and Recognition

  • Cash prizes(in INR)*:
    • Winner: 15K
    • First Runner-up: 10K
    • Second Runner-up: 5K
  • Free registration to top-5 teams in NCVPRIPG'23
  • Opportunity for summer internship in IITJ
  • Paper writing collaboration
  • Certificate to each participant

(*It is mandatory to attend the NCVPRIPG'23 in person in order to be eligible for the prize and certificates.*)


For any queries, please contact Abu Shahid @

Dr. Anand Mishra
Dr. Anand Mishra

Assistant Professor,
Department of CSE,
IIT Jodhpur

Gyan Prabhat
Gyan Prabhat

Ph.D. Student

Prajwal Gatti
Prajwal Gatti

Research Assistant

Abu Shahid
Abu Shahid

Undergraduate Student