Artificial Intelligence


Captcha Generation and Identification Using Generative Adversarial Networks

Authors: Hardik Ajmani, Mrinal Wahal

Adversarial attacking is an emerging worrying angle in the field of AI, capable of fooling even the most efficiently trained models to produce results as and when required. Inversely, the same design powering adversarial attacks can be employed for efficient white-hat modeling of deep neural networks. Recently introduced GANs (Generative Adversarial Networks) serve precisely this purpose by generating forged data. Consequently, authentic data identification is a crucial problem to be done away with, considering increased adversarial attacks. This paper proposes an approach using DCGANs (Deep Convolutional Generative Adversarial Networks) to both - generate and distinguish artificially produced fake captchas. The generator model produces a significant number of unseen images, and the discriminatory model classifies them as fake (0) or genuine (1). Interestingly enough, both the models can be configured to learn from each other and become better as they train along.

Comments: 10 Pages.

Download: PDF

Submission history

[v1] 2019-09-24 23:10:03

Unique-IP document downloads: 44 times is a pre-print repository rather than a journal. Articles hosted may not yet have been verified by peer-review and should be treated as preliminary. In particular, anything that appears to include financial or legal advice or proposed medical treatments should be treated with due caution. will not be responsible for any consequences of actions that result from any form of use of any documents on this website.

Add your own feedback and questions here:
You are equally welcome to be positive or negative about any paper but please be polite. If you are being critical you must mention at least one specific error, otherwise your comment will be deleted as unhelpful.

comments powered by Disqus