Convolutional Neural Network Based on Diverse Gabor Filters for Deepfake Recognition

Abstract

Media synthesis and manipulation has reached unprecedented levels of realism owing to the proliferation of deep learning. Deepfake has been the de-facto tool for media manipulation. Although this technology has potential in the entertainment industry, its threats include political manipulation and bypassing biometric security systems. As a result, deepfake detection has garnered widespread attention among research communities. The intuition is to use deep learning to fix the problems created by deep learning. Although convolutional neural networks have shown their dominance in the filed of pattern recognition, the receptive field-model size dilemma still persists along with the lack of interpretation for such models. While the traditional Gabor function was proposed to fix these problems, it can only generate limited linear Gabor filters which makes it optimal for limited data and applications. The contribution of this paper is quadruple: (i) proposing a unified Gabor function capable of generating linear, elliptical, and circular Gabor filters. (ii) leveraging the back-propagation learning framework to incorporate the proposed function in convolutional neural networks and generate adaptive Gabor filters. (iii) presenting a dual scale large receptive field network for deepfake image recognition. (iv) demonstrating where the proposed model stands in terms of performance and architecture size compared to state-of-the-art models. The proposed model is evaluated on four benchmark datasets: Celeb-DF (v2), DeepFake Detection Challenge Preview, FaceForensics++ and Wilddeepfake. Experimental results show that the proposed adaptive Gabor filters reduces the model size by 64.9% compared to adaptive weighted filters without performance reduction.

A novel forgery classification method based on multi‐scale feature capsule network in mobile edge computing

“…The method uses a capsule network to classify the extracted facial features. 36 Design a nonlinear Gabor filter to fully get the forged features of the face. This method is optimal for dealing with limited face data.…”

“…Nguyen et al 35 find that capsule network contains more spatial relationships of objects than convolutional neural networks. The method uses a capsule network to classify the extracted facial features 36 . Design a nonlinear Gabor filter to fully get the forged features of the face.…”

Deep Fake Video Detection

“…Proposed Accuracy is 77.94%. In this study, we introduced a method named Deep Vision to analyze significant changes in eye blinking, utilizing machine learning, various algorithms, and a heuristic approach to detect Deep fakes generated by GANs.The proposed algorithm, [3] Deep Vision, observed Variations in blinking patterns related to gender, age, and cognitive behavior, employing machine learning techniques. The algorithm, based on previous studies, demonstrated a high accuracy of 87.5% in detecting Deep fakes and normal videos, although acknowledging a limitation related to correlations with mental illness and dopamine activity.…”

Hybrid Biometric Based Person Identification Using Machine Learning

“…The hammering distance is used for recognition. A multi-biometric system based on fingerprints and ear scans [10] has been developed. To start, an adaptive median filter is applied to both biometric images to get rid of unwanted noise.…”