Face Recognition Evaluation for ATM Machines

Face Recognition Evaluation for ATM Machines

Face Recognition Evaluation for ATM Machines

Introduction to face recognition evaluation for ATMs

Automated Teller Machines (ATMs) and other financial transaction kiosks are one of the most popular methods for making financial transactions. Their use of it continues to increase as it’s more convenient for people than visiting a bank and waiting in queues to do the transactions. 

An increase in crimes and fraud attacks

  • The crimes related to ATMs have increased lately as box banking habits increased, and according to the latest ATM Crime Report by the European ATM security team (EAST), the crimes were up 287% in 2017 compared to 2015. 
  • “ATM related fraud attacks increased by 26%, up from 18,738 in 2015 to 23,588 in 2016.  This rise was mainly driven by a 147% increase in Transaction Reversal Fraud (up from 5,104 to 12,581 incidents). The downward trend for card skimming continues with 3,315 card skimming incidents reported, down 20% from 4,131 in 2015. This is the lowest number of skimming incidents reported since 2005.
  • Losses due to ATM-related fraud attacks were up 2% compared with 2015 (up from €327 million to €332 million). Most of such losses were reported in the Asia-Pacific region and the USA. However, domestic skimming losses rose 24% over the same period (up from €44 million to €53 million).
  • ATM related physical attacks rose 12% when compared with 2015 (up from 2,657 to 2,974 incidents).  Within this total, ATM explosive attacks (including explosive gas and solid explosive attacks) were up 47% from the previous year (up from 673 to 988 incidents). Losses due to ATM-related physical attacks were €49 million, unchanged from the previous year.” (URL))

It has been a similar situation in Sri Lanka as well. Even though each Cash Withdrawal / Deposit machines bear the instructions to remove head gears such as caps, helmets, or shades, most people seem to avoid instructions. Despite the security cameras installed at each location, the crimes continue. The need for a more secure method for using ATMs has increased due to such reasons. 

Biometrics to reduce fraudulent attacks

Biometric methods have been extensively used to reduce such crimes and encourage secure banking habits. One of the most popular methods has been used in capturing an image or a video of the user, which later will be used if needed to investigate a criminal act. But, most ATM-based attacks are conducted while shielding the face, which makes it impossible to track the criminals using facial images. Criminals have found tricks to mislead such secure methods, and it increases the vulnerability of that method.

Face recognition

To overcome these problems, this proposal presents a different method of face recognition which will propose how to evaluate face recognition using the exceptional occlusion handling (EOH) method.  Evaluating face recognition using the exceptional occlusion handling (EOH) method will reduce such vulnerabilities and increase the security of Box banking, including ATMs and other transaction kiosks.  

About the Author
The Editorial Team at Netizen Me Insight

The Netizen Me Insight team is a group of passionate writers and editors committed to providing accurate and informative content.

With expertise in areas such as sustainable living, social responsibility, history, art, marketing, business, psychology, and tech, we deliver diverse perspectives and stay up-to-date on the latest developments in our fields.