Face Detection for Secure Online Payment with Proxy Detection

Most of the online transactions use OTP as a way for authenticating the concerned user in addition to their password. The major usage of OTP is to increase security and avoid fraudulent activities such as identity theft. OTP's aren't that secure in some environments. The biggest disadvanta...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:2022 International Conference on Communication, Computing and Internet of Things (IC3IoT) S. 1 - 4
Hauptverfasser: C, Vijayalakshmi, S, Metilda Florence, D, Gokulakrishnan, G, Prakash, Ponmalar, A.
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 10.03.2022
Schlagworte:
Authentication
Classification algorithms
Convolutional neural networks
Detectors
Facenet algorithm
Multi Task Cascaded Convolutional Neural Network
Proxy image
Security
Support Vector Machine
Support vector machines
Telecommunication computing
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Most of the online transactions use OTP as a way for authenticating the concerned user in addition to their password. The major usage of OTP is to increase security and avoid fraudulent activities such as identity theft. OTP's aren't that secure in some environments. The biggest disadvantage of using OTP is that, it could be easily stolen and can be used by a person who has knowledge about the concerned users bank details including access to the registered mobile number, there is a possibility for the amount to be transferred without the account holder's knowledge. To overcome this drawback, we propose a system which uses the Face of the account holder as a way for authentication and verification. The face of the account holder is captured during the initial process. We use FaceNet algorithm to extract facial features. Facial features of every person will be classified using SVM. Face detected during online payment process is compared with the trained face model. If proxy is not detected and the detected face matches the trained face, payment process will be successfully completed.
DOI:10.1109/IC3IOT53935.2022.9767730