# Writer-independent Feature Learning for Offline Signature Verification using Deep Convolutional Neural Networks

arXiv: https://arxiv.org/abs/1604.00974 · HTML: https://mlsanity.com/paper/1604.00974/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 4 Apr 2016.

## Built on by (most cited first, 6 of 6)

- [Detecting Individual Decision-Making Style: Exploring Behavioral Stylometry in Chess](https://mlsanity.com/paper/2208.01366/) (2022)
- [A White-Box False Positive Adversarial Attack Method on Contrastive Loss Based Offline Handwritten Signature Verification Models](https://mlsanity.com/paper/2308.08925/) (2023)
- [A Perspective Analysis of Handwritten Signature Technology](https://mlsanity.com/paper/2405.13555/) (2024)
- [Signature Verification using Geometrical Features and Artificial Neural Network Classifier](https://mlsanity.com/paper/2108.02029/) (2021)
- [Offline Signature Verification Based on Feature Disentangling Aided Variational Autoencoder](https://mlsanity.com/paper/2409.19754/) (2024)
- [Consensus-Threshold Criterion for Offline Signature Verification using Convolutional Neural Network Learned Representations](https://mlsanity.com/paper/2401.03085/) (2024)

## Abstract (arXiv)

Automatic Offline Handwritten Signature Verification has been researched over the last few decades from several perspectives, using insights from graphology, computer vision, signal processing, among others. In spite of the advancements on the field, building classifiers that can separate between genuine signatures and skilled forgeries (forgeries made targeting a particular signature) is still hard. We propose approaching the problem from a feature learning perspective. Our hypothesis is that, in the absence of a good model of the data generation process, it is better to learn the features from data, instead of using hand-crafted features that have no resemblance to the signature generation process. To this end, we use Deep Convolutional Neural Networks to learn features in a writer-independent format, and use this model to obtain a feature representation on another set of users, where we train writer-dependent classifiers. We tested our method in two datasets: GPDS-960 and Brazilian PUC-PR. Our experimental results show that the features learned in a subset of the users are discriminative for the other users, including across different datasets, reaching close to the state-of-the-art in the GPDS dataset, and improving the state-of-the-art in the Brazilian PUC-PR dataset.