XceptionNet-based Digital Image Forensics with DFRWS Framework for Deepfake Detection

Abstrak

This study presents a novel approach to deepfake detection by integrating the DFRWS (Digital Forensics Research Workshop) framework with XceptionNet-based deep learning architecture. The rapid advancement of deepfake technology poses significant threats to digital media authenticity, necessitating robust detection methods. Our research implements a fine-tuned XceptionNet model with additional regularization techniques and focuses on facial feature analysis. The model was trained on a balanced dataset of 2,000 images, equally divided between authentic and deepfake samples. Experimental results demonstrate exceptional performance, achieving 91.25% accuracy, 88.73% precision, 94.50% recall, and an AUC score of 0.9710. The proposed model shows significant improvement in detecting subtle manipulation artifacts while maintaining computational efficiency.

Identifier

• DOI: https://doi.org/10.33506/insect.v11i2.4996
• OpenAlex: https://openalex.org/W4415688863