Deepfake Detection using Integrate-Backward-Integrate Logic Optimization Algorithm with CNN

Deepfake Detection using Integrate-Backward-Integrate Logic Optimization Algorithm with CNN

28/06/2024 | R. Uma Maheshwari, B. Paulchamy, Arun M, Vairaprakash Selvaraj, Dr. N. Naga Saranya and Dr. Sankar Ganesh S
The paper presents a novel approach to deepfake detection by integrating the Integrate-backward-integrate (IbI) Logic Optimization Algorithm with Convolutional Neural Networks (CNNs). The method aims to enhance the CNN's ability to detect subtle nuances between authentic and manipulated visuals through a multi-phase iterative process. Initially, the CNN is trained on a diverse dataset of real and deepfake images. In the forward integration phase, the CNN classifies images as real or deepfake. The backward phase involves the IbI Logic Optimization Algorithm, which uses feedback from the CNN's performance to iteratively refine the network's parameters, architecture, and feature extraction capabilities. This process is repeated through multiple iterations to improve the CNN's accuracy. The re-integration phase evaluates the refined CNN's performance using separate datasets to prevent overfitting and ensure generalizability. The proposed method is designed to adapt to evolving deepfake techniques, addressing the dynamic nature of media manipulation. The effectiveness of the approach is validated through experimental results, showing improved performance metrics compared to existing deepfake detection methods.The paper presents a novel approach to deepfake detection by integrating the Integrate-backward-integrate (IbI) Logic Optimization Algorithm with Convolutional Neural Networks (CNNs). The method aims to enhance the CNN's ability to detect subtle nuances between authentic and manipulated visuals through a multi-phase iterative process. Initially, the CNN is trained on a diverse dataset of real and deepfake images. In the forward integration phase, the CNN classifies images as real or deepfake. The backward phase involves the IbI Logic Optimization Algorithm, which uses feedback from the CNN's performance to iteratively refine the network's parameters, architecture, and feature extraction capabilities. This process is repeated through multiple iterations to improve the CNN's accuracy. The re-integration phase evaluates the refined CNN's performance using separate datasets to prevent overfitting and ensure generalizability. The proposed method is designed to adapt to evolving deepfake techniques, addressing the dynamic nature of media manipulation. The effectiveness of the approach is validated through experimental results, showing improved performance metrics compared to existing deepfake detection methods.
Reach us at info@study.space