Counterfeit Currency Detection using Resource Efficient Neural Networks
Arunabha Mittra1, Indranil Paul2, S Sharanya3
1Arunabha Mittra*, SRM univercity, Kattankulathur, Chennai Tamil Nadu.
2Indranil Paul, SRM Univercity, Kattankulathur, Chennai Tamil Nadu.
3S Sharanya, SRM Univercity, Kattankulathur, Chennai Tamil Nadu.
Manuscript received on March 15, 2020. | Revised Manuscript received on March 27, 2020. | Manuscript published on April 10, 2020. | PP: 220-223 | Volume-9 Issue-6, April 2020. | Retrieval Number: F3626049620/2020©BEIESP | DOI: 10.35940/ijitee.F3626.049620
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: One of the leading causes of economic instability is the large-scale counterfeiting of the paper currency notes. Several media reports bring to light the alarming cases and the humungous scales of currency counterfeiting and how this issue has become very serious now. A report on how the Government is coping with these threats with new and stricter rules however counterfeiters adapt to the new rules in an alarmingly fast pace. Criminals continue to find a loophole in the system despite such strict security features. There have been impressive discoveries in the field of counterfeit currency, and this coupled with new age digital technology, counterfeiting is being fought well. However, it is impossible to track all counterfeit notes and impossible to have them checked at a short amount of time. Existing systems involve filing a case with the police, sending the documents for verification and waiting for the results to come. This method is based on Deep Learning, which has seen tremendous success in image classification tasks in recent times. This technique can help both people and machine in identifying a fake currency note in real time through an image of the same. Traditional Deep Learning algorithms require tremendous amount of compute power and storage and hence it is an expensive and elaborate process. The main goal is to make a faster and simpler mechanism to detect a counterfeit note that can be implemented in any random place like an ATM dispenser or an android application. The success of this application will greatly help the quick identification of the threat and help law enforcement in finding the source of the threat faster.
Keywords: Counterfeit Currency, One-Shot Neural Networks, Deep Learning.
Scope of the Article: Neural Information Processing