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

Abstract

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).

Calibration-Free Monocular Vision-Based Robot Manipulations With Occlusion Awareness

“…Fine-tuning of a model that has been pretrained in ImageNet [26] can give well result in object recognition [27], but it requires a large amount of data for the fine tuning. Meta learning (MAML [28]) is another method that can improve the generalization ability of the model by utilizing a few data from a target domain, which has been widely applied in object detection [29], [30] and classification [31]. Inspired by the mentioned works, this paper proposes to combine real scene data-based fine tuning method with the domain randomization to relax the requirement of designing the simulation parameters to further improve the accuracy of the regular domain randomization.…”

Research advanced in offline handwritten signature verification

“…Kao and Wen [14] proposed an offline signature authenticity detection method based on an explainable deep learning method and a single known sample. Hafemann et al [15] proposed a solution based on meta-learning methods. Since small datasets are more consistent with the application of real practical scenarios, this research area still needs more attention.…”

AVN: An Adversarial Variation Network Model for Handwritten Signature Verification

“…One major stream is deep learning and neural network models [24], [3], [10], [25], [26], [13], which will be detailed in the following section. Other typical methods include meta-learning [27] and one-class method [28]. While these approaches have made impressive progress on signature verification, most of them regard the signification problem as a binary image classification problem and did not specifically consider to mine the information of signature stroke itself but rather the signature image.…”