Signature Embedding: Writer Independent Offline Signature Verification with Deep Metric Learning

Rantzsch, Hannes; Yang, Haojin; Meinel, Christoph

doi:10.1007/978-3-319-50832-0_60

Cited by 31 publications

(24 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this work we consider the two-stage approach described in [5] and [6], where we train Writer-Dependent classifiers on a set of users, using a feature representation learned on another set of users. We note, however, that the methods described in this paper can be used for other feature learning strategies, such as the ones used in [7,8].…”

Section: Proposed Methodsmentioning

confidence: 99%

“…Most of the research effort in offline signature verification has been devoted to finding good feature representations for signatures, by proposing new feature descriptors for the problem [4]. Recent work, however, showed that learning features from data (signature images) can improve system performance to a large extent [5][6][7][8]. These work rely on training Deep Convolutional Neural Networks (CNNs) to learn a hierarchy of representations directly from signature pixels.…”

Section: Introductionmentioning

confidence: 99%

“…The methods proposed in the literature use small images (e.g. 96 × 192 in [7], 170 × 242 in [9]). For the signatures used in these papers, this is equivalently of using a resolution around 100 dpi.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Fixed-sized representation learning from offline handwritten signatures of different sizes

Hafemann

Oliveira

Sabourin

2018

IJDAR

View full text Add to dashboard Cite

This is a pre-print of an article published in the International Journal on Document Analysis and Recognition. The final authenticated version is available online at: https://doi.org/10.1007/s10032-018-0301-6.Abstract Methods for learning feature representations for Offline Handwritten Signature Verification have been successfully proposed in recent literature, using Deep Convolutional Neural Networks to learn representations from signature pixels. Such methods reported large performance improvements compared to handcrafted feature extractors. However, they also introduced an important constraint: the inputs to the neural networks must have a fixed size, while signatures vary significantly in size between different users. In this paper we propose addressing this issue by learning a fixed-sized representation from variable-sized signatures by modifying the network architecture, using Spatial Pyramid Pooling. We also investigate the impact of the resolution of the images used for training, and the impact of adapting (fine-tuning) the representations to new operating conditions (different acquisition protocols, such as writing instruments and scan resolution). On the GPDS dataset, we achieve results comparable with the stateof-the-art, while removing the constraint of having a maximum size for the signatures to be processed. We also show that using higher resolutions (300 or 600dpi) can improve performance when skilled forgeries from a subset of users are available for feature learning, but lower resolutions (around 100dpi) can be used if only L. Hafemann · R. Sabourin Laboratoire d'imagerie, de vision et d'intelligence artificielle (LIVIA) Ecole de technologie supérieure, 1100, rue Notre-Dame Ouest, genuine signatures are used. Lastly, we show that finetuning can improve performance when the operating conditions change.

show abstract

Section: Proposed Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Fixed-sized representation learning from offline handwritten signatures of different sizes

Hafemann

Oliveira

Sabourin

2018

IJDAR

View full text Add to dashboard Cite

show abstract

“…Recent work on signature verification rely on feature learning methods [12], [4], [13], [14], [8], in which learning is conducted directly from signature pixels, instead of relying on handcrafted feature extractors. In this case, a function φ(x) is learned to extract features from signature images x, by training using a surrogate objective, e.g.…”

Section: Related Workmentioning

confidence: 99%

“…Also, these systems use a fixed representation for all users, and it is possible that adapting the representation for each user would yield improvements in classification performance. It is also worth noting that, for WI classification, signature verification systems can be trained jointly (feature extraction and classification) [8]. Despite being jointly trained, such WI systems still perform worse than WD classifiers trained with features learned with surrogate objectives, at least when more than one signature references are used [4].…”

Section: Related Workmentioning

confidence: 99%

Meta-Learning for Fast Classifier Adaptation to New Users of Signature Verification Systems

Hafemann

Sabourin

Oliveira

2020

IEEE Trans.Inform.Forensic Secur.

View full text Add to dashboard Cite

Offline Handwritten Signature verification presents a challenging Pattern Recognition problem, where only knowledge of the positive class is available for training. While classifiers have access to a few genuine signatures for training, during generalization they also need to discriminate forgeries. This is particularly challenging for skilled forgeries, where a forger practices imitating the user's signature, and often is able to create forgeries visually close to the original signatures. Most work in the literature address this issue by training for a surrogate objective: discriminating genuine signatures of a user and random forgeries (signatures from other users). In this work, we propose a solution for this problem based on meta-learning, where there are two levels of learning: a task-level (where a task is to learn a classifier for a given user) and a meta-level (learning across tasks). In particular, the meta-learner guides the adaptation (learning) of a classifier for each user, which is a lightweight operation that only requires genuine signatures. The meta-learning procedure learns what is common for the classification across different users. In a scenario where skilled forgeries from a subset of users are available, the meta-learner can guide classifiers to be discriminative of skilled forgeries even if the classifiers themselves do not use skilled forgeries for learning. Experiments conducted on the GPDS-960 dataset show improved performance compared to Writer-Independent systems, and achieve results comparable to state-of-the-art Writer-Dependent systems in the regime of few samples per user (5 reference signatures).

show abstract

Offline Signature Verification Using Textural Descriptors

Hadjadj

Gattal

Djeddi

et al. 2019

Pattern Recognition and Image Analysis

View full text Add to dashboard Cite

Signature Embedding: Writer Independent Offline Signature Verification with Deep Metric Learning

Cited by 31 publications

References 12 publications

Fixed-sized representation learning from offline handwritten signatures of different sizes

Fixed-sized representation learning from offline handwritten signatures of different sizes

Meta-Learning for Fast Classifier Adaptation to New Users of Signature Verification Systems

Offline Signature Verification Using Textural Descriptors

Contact Info

Product

Resources

About