• NAINESH GATHIYAWALA
• Dr. D.B.Shah

In the ever-changing terrain of artificial intelligence and behavioral biometrics, with much cross-pollination between psychology and machine learning, new possibilities emerge for non-invasive personality assessments. It introduces an intelligent system to predict personality traits through signature analysis using deep learning. Signatures are unique to individuals and serve as legal biometric proof; they represent the fine gradations of human behavior, writing pressure, slant, loop, and flow, which are personality markers acknowledged in graphology and cognitive science. This model has as its heart a custom-designed Convolutional Neural Network trained with 601 annotated real-world images of signatures, each annotated with one of eight personality traits, such as Assertive, Introverted, Creative, Analytical, etc. Extensive experiments showcased the custom CNN architecture as being superior to pre-trained architectures such as VGG16 (93% accuracy), ResNet (91.36%), and YOLO (89.92%), and attaining a classification accuracy of 97.31%. This CNN model used successively deep convolution layers followed by pooling and dropout, making it able to extract hierarchical features from complex patterns and reduce overfitting. The research is carried out as a full-stack implementation, to ensure that it is truly practical. A user-friendly interface is built via the Web for uploading signature images that would be processed through a backend based on FastAPI. The backend loads the trained model for performing image preprocessing operations, such as resizing, normalization, and augmentation, and returns predictions. The frontend also provides visual feedback with both the uploaded image and the predicted personality. This model in itself offers plenty of applications: assessing behavioral compatibility in recruitment, criminal detection, and the victims of fraud in forensics, counseling ideas, human-computer interface for adaptive personalization-The idea of combining social theories with modern deep-learning models represents a move from classic behavioral analysis to computation that is efficient and objective-witnessing a scalable path toward automatic personality classification.

signature Analysis,graphology,cnn model,personality traits

"A Human Signature-Based Model for Behavioral Trait Analysis", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.13, Issue 1, page no.a788-a800, January-2026, Available :http://www.jetir.org/papers/JETIR2601093.pdf

"A Human Signature-Based Model for Behavioral Trait Analysis", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.13, Issue 1, page no. ppa788-a800, January-2026, Available at : http://www.jetir.org/papers/JETIR2601093.pdf