Identifying Fake Images through CNN Based Classifaction Using FIDAC
→
Summary
Designed and implemented a Convolutional Neural Network (CNN) model to classify real vs. fake images. Integrated the FIDAC (Fake Image Detection and Classification) framework to enhance detection accuracy. Preprocessed datasets with normalization and augmentation to improve model generalization. Extracted deep visual features using convolution and pooling layers for robust classification. Evaluated performance using metrics such as accuracy, precision, recall, and F1-score. Achieved strong resilience against manipulation techniques like splicing, copy-move, and deepfakes. Delivered a scalable solution applicable to digital forensics, media verification, and security domains. Contributed to combating misinformation and image-based fraud by ensuring authenticity in digital content.