FAKE IMAGE DETECTION USING DEEP LEARNING TECHNIQUES

This project presents a deep learning-based approach for detecting fake or tampered images. A total of 3000
images were used to train and evaluate the model, consisting of both real and manipulated images. The
dataset was divided into training and testing sets to ensure proper model evaluation. Error Level Analysis
(ELA) was applied to highlight potential manipulated regions before feeding the images into a
Convolutional Neural Network (CNN) for classification. Preprocessing techniques such as resizing,
normalization, and data augmentation were implemented to improve performance and generalization. The
trained model achieved an accuracy of 96% on the testing dataset, demonstrating its effectiveness in
identifying digitally altered images.