Recurrent Binary Patterns and CNNs for Offline Signature Verification

Yılmaz, Mustafa Berkay; Ozturk, Kagan

doi:10.1007/978-3-030-32523-7_29

Cited by 6 publications

(5 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In work [20], a two-channel CNN is used for feature extraction and introduces additional genuine signatures in the model for reference. In work [40], the feature extraction is performed with the proposed Recurrent Binary Patterns (RBP) network combined with the two-channel CNN which was proposed in work [20]. Work [43] extracts four kinds of features and constructs classifiers by using RNNs.…”

Section: Experimental Results On the Gpds Datasetmentioning

confidence: 99%

Deep Canonically Correlated Denoising Autoencoders for Signature Verification Purpose

Zhao,

Chen,

Yuan

et al. 2024

Preprint

View full text Add to dashboard Cite

With the rapid development of modern Internet of Things (IoT) technology, thehandwritten signature verification system becomes a typical Human-ComputerInteraction (HCI) application, which is often applied to many authorization documents with legal uses. To build a high-performance verification system, featureextraction is one of the most crucial steps. However, the generalization abilityof traditional deep learning-based feature extractors is not always satisfactorysince most feature extractors are only suitable for source signature datasets.In this paper, to improve the generalization ability of existing deep learning-based feature extractors for the offline signature verification task, we proposea novel multi-view learning-based framework, named Deep Canonically Corre-lated Denoising Autoencoders (DCCDAE). Specifically, the DCCDAE generatesthe joint features as the final features based on original deep learning-basedfeatures and another noisy view of them by optimizing the Canonically Cor-related Analysis (CCA) objective and minimizing the reconstruction error oforiginal features. Extensive experiments and discussions on four publicly avail-able datasets, GPDS, CEDAR, MCYT-75, and PUC-PR demonstrate that theproposed DCCDAE can improve the generalization ability of the original deeplearning-based feature extractor and achieve the state-of-the-art performancecompared with other offline signature verification systems. The code is availableat https://github.com/star0511/DCCDAE

show abstract

Section: Experimental Results On the Gpds Datasetmentioning

confidence: 99%

Deep Canonically Correlated Denoising Autoencoders for Signature Verification Purpose

Zhao,

Chen,

Yuan

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…For any given population of the ω L+ (3G) samples we solve eqs. (9,10) for evaluation of the dispersion parameter Table 1 provides a summary of the proposed experimental setups. The ω T+ class contains the remaining genuine samples ω T+ (G)…”

Section: Methods and Experimental Resultsmentioning

confidence: 99%

“…signal vs. time) [5][6][7] or static-offline (i.e. image) [8][9][10][11][12][13]. An alternative classification of offline signature verification methodologies divides them into a) handcrafted methods, which mainly utilize image processing and computer vision techniques and b) data-driven or learningbased approaches with typical representatives Bags of Visual Words [14,15] sparse representation [11] and deep learning methodologies [8,12,[16][17][18][19][20][21][22][23][24][25][26].…”

Section: Introductionmentioning

confidence: 99%

“…The resulting network is contained within the feature extraction stage providing a vectored representation for any new test signature image [12,19,31,32]. A number of deep learning based topologies are obtained by examining different loss functions [31] and similarity strategies [8,10,30] commonly used on the dominant SigNet architecture. Also, post feature management methods are applied, exploiting the effectiveness of the extracted vectors [13,33].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Subscripto multiplex: A Riemannian symmetric positive definite strategy for offline signature verification

Zois

Said

Tsourounis

et al. 2023

Pattern Recognition Letters

View full text Add to dashboard Cite

“…While previous research efforts have made great progress over the past decades [1], [2], [3], [11], [12], [13], [14], [15], [16], [17], there is still much space to address the underlying essential challenges and improve the signature verification performance. The early research efforts extract hand-crafted features and compare the distance between two signatures for verification, such as signature heights, widths, areas [1], [18], local patches LBP [19], [20], [21], [22], and SIFT feature [11].…”

Section: A Signature Verificationmentioning

confidence: 99%

AVN: An Adversarial Variation Network Model for Handwritten Signature Verification

Wei

2022

IEEE Trans. Multimedia

View full text Add to dashboard Cite

Handwritten signature verification is a crucial yet challenging problem. While previous studies have made great progress in this problem, they learn signature features passively from given existing data. In this paper, we propose a novel adversarial variation network (AVN) model for handwritten signature verification which mines effective features by actively varying existing data and generating new data. Powered by a proposed novel variation consistency mechanism, the AVN contains three different types of modules unified under one end-to-end framework: the extractor seeks to extract deep discriminative features of handwritten signatures, the discriminator aims to make verification decisions based on the extracted features, and the variator is designed to actively generate signature variants for constructing a more discriminative model. The proposed model is trained in an adversarial way with a min-max loss function, by which the three modules cooperate and compete to enhance the entire model's ability and therefore the signature verification performance is improved. We test the proposed method on four challenging signature datasets of different languages: CEDAR, BHSig-Hindi, BHSig-Bengali, and GPDS Synthetic Signature. Extensive experiments with in-depth discussions demonstrate the strength and potential of the proposed method.

show abstract

Recurrent Binary Patterns and CNNs for Offline Signature Verification

Cited by 6 publications

References 29 publications

Deep Canonically Correlated Denoising Autoencoders for Signature Verification Purpose

Deep Canonically Correlated Denoising Autoencoders for Signature Verification Purpose

Subscripto multiplex: A Riemannian symmetric positive definite strategy for offline signature verification

AVN: An Adversarial Variation Network Model for Handwritten Signature Verification

Contact Info

Product

Resources

About