Aghera Punamben J.
Computer Science
March 2025
Handwritten signatures embody different informational characteristics that make them good markers for personality traits. The research work aims at predicting the Big Five personality traits of Neuroticism, Agreeableness, Extraversion, Conscientiousness, and Openness with the aid of the advanced deep learning techniques used for handwritten signature feature extraction. The public ally accessible dataset of pre-annotated signatures with personality labels is modelled for black and white conversion, Gaussian noise filtering, image normalization, and binarization to conduct more effective feature extraction. Convolutional Neural Networks (CNNs) learn the important features of a signature without human intervention; parameters such as stroke width, slant angle, pressure variation, and signature length are learned with ResNet and VGG architectures
300- 316
© Airo Journals 2021
