Manuscript Title:

ATM FRAUD DETECTION USING DEEP LEARNING

Author:

SHAIK DILSHATH, SANJANA ATHMARAMAN, PRASANNA LAKSHMI, Prof. SIVA SHANMUAGM G

DOI Number:

DOI:10.17605/OSF.IO/M3YTH

Published : 2021-11-23

About the author(s)

1. SHAIK DILSHATH - Department of Computer Science Engineering, Vellore Institute of Technology, Vellore, Tamil Nadu, India
2. SANJANA ATHMARAMAN - Department of Computer Science Engineering, Vellore Institute of Technology, Vellore, Tamil Nadu, India.
3. PRASANNA LAKSHMI - Department of Computer Science Engineering, Vellore Institute of Technology, Vellore, Tamil Nadu, India
4. SIVA SHANMUAGM G - Department of Computer Science Engineering, Vellore Institute of Technology, Vellore, Tamil Nadu, India.

Full Text : PDF

Abstract

An Automatic Teller Machine is an automated electronic bank machine that helps the customer with monetary transitions without any help of bank officials or intermediaries. Customers can access cash transitions and details through the machine, which is why ATMs carry large amounts of cash, thus prone to attacks. Physical attacks can be attempted in the ATM, by means of thermal or mechanical ways with the goal to break or harm the ATM machine so as to rob the machine by stealing the cash inside it. The techniques which are generally used are explosive attacks, ram-raids and cutting. The money inside the ATM can be stolen when it is being repaired or when the money is being stored. Staff is either held up as they are conveying cash to or from an ATM or when the ATM safe is open and money tapes supplanted. There is an assortment of physical and mechanical components that can restrain assaults to the safe.Thus, in this project, we tried to predict if the attack is happening and thus send the message to the local police station. This project helps in predicting the attack before it has already occurred.


Keywords

Camera, internet connection, CNN, Fast RCNN, FaceNet, haar cascades.