A Light Weight and Hybrid Deep Learning Model Based Online Signature Verification

The augmented usage of deep learning-based models for various AI related problems are as a result of modern architectures of deeper length and the availability of voluminous interpreted datasets. The models based on these architectures require huge training and storage cost, which makes them inefficient to use in real time systems like online signature verification (OSV) and to use in resource restraint devices. As a solution, in this work, our contribution is two-fold. 1) An efficient dimensionality reduction technique, to lessen the number of features to be considered and 2) a state-of-the-art model CNN-LSTM based hybrid architecture for online signature verification. Thorough experiments on the openly accessible datasets MCYT, SUSIG, SVC approves that the proposed framework attains improved precision even with as low as one training sample. The proposed framework produce state-of-the-art performance in various categories of all the three datasets.