Attention-enhanced wavelet high-frequency residuals for efficient privacy-preserving face imaging and recognition

Wavelet-based features to improve cross-forgery generalization in deepfake detection

Deepfake detection remains a challenging problem due to the ever-evolving techniques used to synthesize realistic fake images and videos. Traditional detection methods primarily operate in the pixel domain, often struggling to generalize across different forgery types. In this work, we explore an approach that integrates wavelet-based frequency analysis with deep learning to enhance deepfake detection. By decomposing input images using the Discrete Wavelet Transform (DWT), we extract multi-resolution features that capture both global structures and fine-grained manipulation artifacts; such features are then input to a neural network. Our results show that level-1 DWT features mainly induce a sensitivity-specificity tradeoff: they improve fake-class sensitivity and F1 in cross-forgery settings, while RGB models retain stronger specificity and in-forgery performance. This suggests that wavelet sub-bands expose manipulation cues that are useful for detecting unseen fake classes, but they should not be interpreted as a uniform robustness improvement.

Application of Image Enhancement Techniques in Facial Recognition System

Face recognition becomes an important biometric implementation in any surveillance, access control, forensic and intelligent security system applications. Despite of technical advances in deep learning models, recognition accuracy normally affected by facial pose, illumination, occlusion, expression, aging factors and limited availability of labelled dataset. These challenges reduce the robustness of face recognition models in real world environments [1][2][3]. Recent researches demonstrated that the super-resolution techniques based on Generative Adversarial Networks (GANs) reconstruct high-quality facial images very effectively from low-resolution input images. It supports to improve feature representation and face recognition [4][5]. Therefore, this research motivated from this advancement and proposes an Adaptive Super-Resolution Generative Adversarial Network (Adaptive SRGAN) for face recognition. It integrates adaptive learning with image super resolution to reconstruct identity preserving high resolution facial images by employing adaptive learning rate optimization, dynamic loss weighting, attention guided feature enhancement and identity preserving loss functions.  However, it enhances reconstruction quality by preserving discriminative facial characteristics [6]. The proposed model is expected to achieve higher Peak Signal to Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), recognition accuracy, precision, recall, and F1-score while reducing false acceptance and false rejection rates. Simultaneously, Adaptive SRGAN delivers a robust and scalable solution for improving face recognition systems.

Deepfake Detection via Frequency-Aware Vision Transformer and Bidirectional Cross-Attention Fusion with Post-Processing Robustness

Today, the use of increasingly ubiquitous synthetic media, or ‘deepfakes’, has become a risk to online trust, information integrity and individual security and is being created by artificial intelligence (AI). The current approaches are mainly based on either spatial features of CNNs or high-level semantic representations of Vision Transformer; both have major drawbacks in effectively leveraging multi-domain forensic cues. This paper presents FAViT (Frequency-Aware Vision Transformer), a hybrid architecture capable of jointly utilizing spatial- and frequency-domain forensic information by the means of a bidirectional cross-attention fusion scheme. We use an 11-channel forensic tensor in each face image (including per-channel Fast Fourier Transform (FFT) magnitude maps, Discrete Wavelet Transform (DWT) sub-bands, channel noise residual maps, Sobel gradient magnitude and channels of Error Level Analysis (ELA)). A Frequency Branch CNN processes this multi-domain tensor and the original RGB image is encoded with a pretrained ViT-B/16 spatial branch. The two streams are combined through the bidirectional cross-attention which allows the model to localize both spatial and spectral manipulation artifacts. We also present an adversarial cleaning simulation pipeline which partitions the training process with five post-processing attack methods, namely GFPGAN neural face restoration, learned autoencoder cleaning, etc., to increase resistance to real-world forensic defenses. Tests of FaceForensics++ C23 (7926 images, consisting of four manipulation types) show that FAViT attains F1-score of 86.22, AUC-ROC of 94.26 and accuracy of 85.55 on the held-out test set. The strength analysis of 21 attack conditions shows that the max degradation in AUC is 30.3, with specific strengths in GFPGAN restoration (AUC = 98.51). Robustness is evaluated based on 21 post-processing attack cases that include JPEG compression, Gaussian blurring, down-sampling, and GFDGAN neural-based restoration; it should be noted that robustness against gradient-based adaptive attacks requires additional attention. Testing on the CIFAKE and Celeb-DF v2 datasets reveals some limitations of domain generalization.

SRAP: SVD-Refined Adversarial Perturbations for Imperceptible Face-Swap Defense

This work proposes SRAP, which combines per-channel truncated SVD refinement with an identity-importance mask at every optimization step, and demonstrates that SRAP substantially improves protected-image fidelity across all reported metrics while maintaining competitive identity-disruption performance.

JFDFusion: joint frequency decomposition and deep feature learning for infrared–visible image fusion

Enhancing face recognition privacy through the integration of differential privacy and convolutional neural network

The research offers an in-depth analysis of various DP techniques to construct a secure face recognition system employing a Convolutional Neural Network and face classifiers, and concludes that the DP blur with Logistic regression predictors provides the highest privacy, achieving excellent accuracy rates of 97% and 77% for these datasets.