Deepfakes, synthetic media created using advanced machine learning techniques, pose significant societal challenges by spreading misinformation and undermining trust in media. With the increasing sophistication of deepfake technologies, distinguishing between genuine and synthetic media has become increasingly difficult. This paper presents a robust deepfake image detection framework using the Swin-B Transformer, a pre-trained model fine-tuned for our application. By integrating a hybrid dataset that combines real images from the FFHQ dataset and synthetically generated fake images from a publicly available Kaggle dataset, we simulate real-world media scenarios. Our model achieves an impressive accuracy of 97.47\% on the test set, demonstrating superior generalization to both real and synthetic visual data. Using Grad-CAM, we visualize the spatial segments of the image that the model focuses on during classification, providing insight into the decision-making process. This work contributes to enhancing content authenticity, controlling fake news, and ensuring digital trust and safety.
Real-time object detection needs to strike a balance between accuracy, speed, and reliability. However, many YOLO-based models struggle with two main issues: weakened feature quality in deeper backbone layers and inefficient multiscale fusion in the detection neck. To overcome these challenges, we present EnYOLO-World — a refined detection framework that combines the YOLOv9 GELAN backbone with Programmable Gradient Information (PGI) to retain detailed spatial features and ensure smoother gradi- ent flow. We also introduce a FiLM-Driven Path Aggregation Network (FiLM-PAN), which adaptively fuses multiscale features using channel-wise affine transformations, achieving efficient performance with linear complexity. Tests on the COCO dataset show that EnYOLO-World offers a better trade-off between accuracy and speed compared to existing real-time detectors, demonstrating the value of optimizing both backbone feature retention and neck-level fusion strategies.
Quantitative image quality assessment (IQA) is a central task in computer vision. This work presents a unified framework, based on directional curvature analysis, that can be parameterized to operate in multiple functional modes. It was demonstrated that the framework can be optimized to act as a state-of-the-art no-reference quality metric (NR-IQA) for two distinct distortion classes. Gaussian blur (p = 0.933) and white noise (p = -0.948), in correlation with human perception in the LIVE dataset. The main contribution is the formulation of a hybrid system that uses the response signature of two optimized filters to first diagnose the artifact type (blur vs. noise) with over 97% accuracy and, then, quantify the magnitude of the degradation using an expert regression model. The final hybrid system achieved predictive performance with an RMSE of 5.29 DMOS points on the combined dataset, establishing it as a robust tool for image quality diagnosis.
Parkinson's disease (PD) is a progressive neurodegenerative disorder that leads to patients’ motor and non-motor disabilities, making early detection challenging. One emerging method for diagnosing PD involves analyzing hand-drawing patterns, such as drawing spirals, circles, or specific shapes. Recent advances in computational techniques, particularly machine learning and deep learning algorithms, have been applied to quantify and assess these drawings, allowing for the differentiation between healthy individuals and those with early-stage PD. In this paper, we propose a multi-modal deep learning model for Parkinson’s disease detection using circles, spirals, and meanders drawings from the NewHandPD database. The proposed approach involves three key stages. First, a dedicated deep learning model is assigned to the three modalities to extract features from the hand-drawn data. These extracted features are then concatenated and fed into a fully connected layer, followed by a softmax classifier for final prediction. To further enhance detection accuracy, the Binary Grey Wolf Optimizer is integrated into the training process to perform feature selection. This method evaluates different combinations of features during training and selects the best subset based on the accuracy maximization. Simulation results using the NewHandPD dataset demonstrated an improvement in Parkinson's disease detection through multimodal and multi-deep learning classification. Furthermore, the introduction of feature selection significantly improved accuracy and reduced the feature vector length.
Gait recognition has emerged as an important biometric modality due to its non-invasive nature and suitability for surveillance and security applications. However, achieving robustness under real-world variations in viewpoint, clothing, and carrying conditions remains a significant challenge. This paper introduces \textbf{AttIncGait}, a deep learning framework that integrates Inception-based multi-scale feature extraction with dual-path attention for effective gait recognition. Unlike prior methods that treat attention as an auxiliary or late-stage refinement, our approach embeds spatial and channel attention directly within Inception modules, enabling simultaneous multi-scale representation and adaptive relevance weighting. This structural integration enhances discriminative capability while preserving computational efficiency. Experiments on CASIA-B and OU-MVLP datasets demonstrate state-of-the-art performance: 97.5% accuracy on OU-MVLP and a 2.6% improvement over the best existing method under clothing variation in CASIA-B. Ablation studies further reveal that spatial and channel attention individually improve accuracy, while their joint integration yields an overall +8.5% gain on OU-MVLP. These results validate the effectiveness of attention-driven multi-scale fusion for gait recognition and highlight the potential of AttIncGait for real-world biometric identification and mobility analysis.
