| Triple Modal Signature (TMS) Dataset |
Features 70 writers, each providing 10 genuine signatures; 20 writers also contributed 10 forged signatures. |
2024 |
Access Link |
Cite |
Zhaosen Shi |
| Offline Handwriting Signature Dataset |
Contains 12,600 offline signature images from 420 individuals, with 30 signatures per participant. |
2023 |
Access Link |
Cite |
University of Raparin and Firat University |
| ChiSig |
A Chinese document signature forgery detection benchmark containing 10,242 images across 500 distinct signed names. |
2023 |
Access Link |
Cite |
Zhejiang University |
| MSDS Dataset |
It has two subsets, and each subset includes 16,080 samples from 402 users, with 20 genuine and 20 skilled forgery samples per user. |
2022 |
Access Link |
Cite |
HCIILAB |
| BHSig260 |
Comprising 260 signers, the dataset includes 14,040 signature images. |
2014 |
Access Link |
Cite |
German Research Center for AI |
| ICDAR 2013 |
The ICDAR 2013 dataset, published by the International Conference on Document Analysis and Recognition (ICDAR) in 2013. |
2013 |
Access Link |
Cite |
ICDAR |
| ICDAR 2011 |
It is an authoritative benchmark dataset released by the International Conference on Document Analysis and Recognition (ICDAR). |
2011 |
Access Link |
Cite |
ICDAR |
| SVC2004 |
This dataset consists of 4,000 signature samples from 100 authors. |
2004 |
Access Link |
Cite |
HKUST |
| MCYT-SignatureOff-75 |
The dataset consists of handwritten samples from 75 signers, totaling 2250 signature images. |
2004 |
Access Link |
Cite |
Autonomous University of Madrid |
| MCYT-Signature-100 |
The database comprises handwritten signature data from 100 users, totaling 5,000 samples. |
2003 |
Access Link |
Cite |
Autonomous University of Madrid |