Tempered Image Detection Using ELA and Convolutional Neural Networks
Anupama Mishra, Kwok Tai Chui, Brij Bhooshan Gupta · 2024
Images are often manipulated to benefit one party, serving as crucial evidence. This manipulation, often used in fake news or misleading information, frequently involves image falsification. Detecting such falsification requires a robust model capable of processing vast image data efficiently. In today’s data-rich era, Deep Learning, especially the use of Convolutional Neural Networks with Error Level Analysis, has achieved an impressive 87.75% accuracy and convergence in 10 epochs for detecting forged images.