Copyright © 2025 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{182363,
        author = {Vinay Gurugubelli and Prashant Kulkarni and Shubhangi Tidake},
        title = {Opportunities and challenges of Gen Ai in Banking sector.},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {2},
        pages = {2025-2031},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=182363},
        abstract = {This work studies the efficiency of deep advanced learning models on detecting financial fraudulent activities in synthetic banking data. A comparative study is executed using ViT (Vision Transformer), CNN, and a hybrid of FPN + PANet to detect fraudulent activities. The preprocessing procedures for the data include feature scaling, label encoding, and conversion of tabular data to image formats. The dataset is extremely imbalanced, thereby presenting problems in finding rare cases of fraud. Accuracy, precision, recall, and F1-score are used as evaluation criteria for judging the performance of the models. The models had a very high accuracy of about 99%, yet the classification report recorded a low recall rate for fraudulent transactions, considering them a big drawback in their ability to handle imbalanced data.},
        keywords = {Financial fraud detection, CNN, Vision Transformer, FPN + PANet, deep learning, class imbalance, synthetic transaction dataset, binary classification, anomaly detection, banking AI.},
        month = {July},
        }