SCRS Conference Proceedings on Intelligent Systems
Image Forgery Detection Using CNN
Authors: Aadeesh Jain, Aditya Sharma, Kanishk Gupta, Ketan Likhi, Neha Mehra, Sonika Shrivastava and Divyansh Joshi
Publishing Date: 25-04-2022
ISBN: 978-93-91842-08-6
Abstract
In today’s world, digital images are widely used in various domains such as; newspapers, scientific journals, magazines, and many other fields. Unfortunately, today’s digital technology made it easy for digital images to be forged due to the availability of the low cost photo editing software like Adobe Photoshop. Thus, in order to recover people’s trust towards digital images, it is important to develop new trustworthy techniques for digital images forgery detection. In this paper we present a novel fake image detection model where the acquisition method is based on an in-depth, process-based, convolutional neural network (CNN) for automatic learning and identifying independent presentations from color RGB input images. CNN is being used because of the advantages it provides as it performs feature engineering, i.e. feature extraction and feature selection, which was earlier performed using different statistical observations and methods. The research paper discusses transfer learning approach which has its own advantages, as it's a different approach from the custom CNN which was being used in earlier approaches.
Keywords
Deep Learning, CNN, Image Processing, Transfer Learning.
Cite as
Aadeesh Jain, Aditya Sharma, Kanishk Gupta, Ketan Likhi, Neha Mehra, Sonika Shrivastava and Divyansh Joshi, "Image Forgery Detection Using CNN", In: Raju Pal and Praveen Kumar Shukla (eds), SCRS Conference Proceedings on Intelligent Systems, SCRS, India, 2022, pp. 293-304. https://doi.org/10.52458/978-93-91842-08-6-29