Characterization of Handwritten Signature Images in Dissimilarity Representation Space

Souza, Victor L. F.; Oliveira, Adriano L. I.; Cruz, Rafael M. O.; Sabourin, Robert

doi:10.1007/978-3-030-22744-9_15

Cited by 3 publications

(8 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The dichotomy transformation is a very important technique to solve some of the problems related to HSV problem, as demonstrated in two preliminary studies (Souza et al, 2019b) and (Souza et al, 2019a). In (Souza et al, 2019b), we have shown that even DT increasing the number of samples in the offline WI-HSV context, as a consequence, redundant information is generated.…”

Section: Introductionmentioning

confidence: 80%

“…For the instance hardness (IH) analysis, K = 7 is used for the estimation of the kDN (Cruz et al, 2017). For the comparative analysis, ten replications were carried out, as in (Souza et al, 2019a).…”

Section: Methodsmentioning

confidence: 99%

“…Also, by increasing the number of samples, the model may be able to obtain sufficient information to capture the full range of signature variations, reducing the effects of the intra-class variability (Hafemann et al, 2017b) However, many of the samples generated by DT in the WI-HSV scenario represent redundant information and therefore have little importance for training purposes. In (Souza et al, 2019b) and (Souza et al, 2019a) we showed that using prototype selection (PS) techniques, such as the Condensed Nearest Neighbors (CNN) (Hart, 1968), in DS, allowed the WI classifier to be trained in the preprocessed DS without deteriorating the verification performance.…”

Section: Dsmentioning

confidence: 99%

“…1. The use of a methodology for analysing the instance hardness (IH), considering the neighborhood of the training set of the GPDS dataset (instead of considering the test sets, as in Souza et al (2019b) and Souza et al (2019a)); 2. The characterization of the dynamic reference selection using the MAX as fusion function;…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A white-box analysis on the writer-independent dichotomy transformation applied to offline handwritten signature verification

Souza

Oliveira

Cruz

et al. 2020

Expert Systems with Applications

Self Cite

View full text Add to dashboard Cite

High number of writers, small number of training samples per writer with high intra-class variability and heavily imbalanced class distributions are among the challenges and difficulties of the offline Handwritten Signature Verification (HSV) problem. A good alternative to tackle these issues is to use a writerindependent (WI) framework. In WI systems, a single model is trained to perform signature verification for all writers from a dissimilarity space generated by the dichotomy transformation. Among the advantages of this framework is its scalability to deal with some of these challenges and its ease in managing new writers, and hence of being used in a transfer learning context. In this work, we present a white-box analysis of this approach highlighting how it handles the challenges, the dynamic selection of references through fusion function, and its application for transfer learning. All the analyses are carried out at the instance level using the instance hardness (IH) measure. The experimental results show that, using the IH analysis, we were able to characterize "good" and "bad" quality skilled forgeries as well as the frontier region between positive and negative samples. This enables futures investigations on methods for improving discrimination between genuine signatures and skilled forgeries by considering * Corresponding author

show abstract

Section: Introductionmentioning

confidence: 80%

Section: Methodsmentioning

confidence: 99%

Section: Dsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A white-box analysis on the writer-independent dichotomy transformation applied to offline handwritten signature verification

Souza

Oliveira

Cruz

et al. 2020

Expert Systems with Applications

Self Cite

View full text Add to dashboard Cite

show abstract

“…A total of 12 references per writer were considered. The DT generates a different dissimilarity vector from each of these references which are passed down to the WI classifier and then the MAX of the signed distance is used as the partial decision fusion function [18]. In the training step (training and validation sets), the model uses genuine signatures and random forgeries.…”

Section: B Experimental Setupmentioning

confidence: 99%

An Investigation of Feature Selection and Transfer Learning for Writer-Independent Offline Handwritten Signature Verification

Souza¹,

Oliveira²,

Cruz³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

An Investigation of Feature Selection and Transfer Learning for Writer-Independent Offline Handwritten Signature Verification

Souza

Oliveira

Cruz

et al. 2021

2020 25th International Conference on Pattern Recognition (ICPR)

View full text Add to dashboard Cite

Characterization of Handwritten Signature Images in Dissimilarity Representation Space

Cited by 3 publications

References 14 publications

A white-box analysis on the writer-independent dichotomy transformation applied to offline handwritten signature verification

A white-box analysis on the writer-independent dichotomy transformation applied to offline handwritten signature verification

An Investigation of Feature Selection and Transfer Learning for Writer-Independent Offline Handwritten Signature Verification

An Investigation of Feature Selection and Transfer Learning for Writer-Independent Offline Handwritten Signature Verification

Contact Info

Product

Resources

About