Explore the vast range of titles available.

Thousands of videos are posted on websites and social media every day, including Twitter, Facebook, WhatsApp, Instagram, and YouTube. Newspapers, law enforcement publications, criminal investigations, surveillance systems, Banking, the museum, the military, imaging in medicine, insurance claims, and consumer photography are just a few examples of places where important visual data may be obtained. Thus, the emergence of powerful processing tools that can be easily made available online poses a huge threat to the authenticity of videos. Therefore, it’s vital to distinguish between true and fake data. Digital video forgery detection techniques are used to validate and check the realness of digital video content. Deep learning algorithms lately sparked a lot of interest in the field of digital forensics, such as Recurrent Neural Networks (RNN), Deep Convolutional Neural Networks (DCNN), and Adaptive Neural Networks (ANN). In this paper, we give a soft taxonomy as well as a thorough overview of recent research on multimedia falsification detection systems. First, the basic knowledge needed to comprehend video forgery is provided. Then, a summary of active and passive video manipulation detection approaches is provided. Anti-forensics, compression video methods, datasets required for video forensics, and challenges of video detection approaches are also addressed. Following that, we presented an overview of deepfake, and the datasets required for detection were also provided. Also, helpful software packages and forensics tools for video detection are covered. In addition, this paper provides an overview of video analysis tools that are used in video forensic applications. Finally, we highlight research difficulties as well as interesting research avenues. In short, this survey provides detailed information and a broader investigation to extract data and detect fraud video contents under one umbrella.

The non-orthogonal multiple access (NOMA) system is considered an important technology that enables the fifth-generation (5G) wireless systems and beyond to satisfy different requirements such as high efficiency, massive networks, sophisticated optimization, and steady quality. Also, the NOMA scheme provides advancements and attractive characteristics such as low latency, ultra-dense service, great fairness, innovative waveform architecture, efficient bandwidth utilization, and massive device connectivity compared to the earliest multiple access schemes. Therefore, the NOMA system necessitates an efficient resource allocation technique such as user pairing (UP) and power allocation (PA) schemes to achieve optimal performance. So, in this paper, we discuss the significance of resource allocation in NOMA in 5G networks and beyond in-depth. As a result, firstly, we review the classification of multiple access schemes, the various types of NOMA techniques, and the characteristics of NOMA. Then, the paper analyzes the issue of resource allocation by classifying the different resource allocation schemes in 5G. Further, a summary of the solutions to the current resource allocation issues are reviewed. Finally, we suggest future research challenges on which to focus.

There are several factors for vehicle accidents during driving such as drivers’ negligence, drowsiness, and fatigue. These accidents can be avoided, if drivers are warned in time. Moreover, recent developments in computer vision and artificial intelligence (AI) have helped to monitor drivers and alert them in case they are not concentrating on driving. The AI techniques can extract relevant features from expressions of driver’s face, such as eye closure, yawning, and head movements to infer the level of sleepiness. In addition, they can acquire biological signals from the driver’s body, and indications from the vehicle behavior. This paper provides a comprehensive review of the detection techniques of drowsiness and fatigue of drivers using machine learning (ML) and deep learning (DL). The current techniques for this application are classified into four categories: image- or video-based analysis during the driving, biological signal analysis for drivers, vehicle movement analysis, and hybrid techniques. A review of supervised techniques is presented for detecting fatigue and drowsiness on different datasets, with a comparison of the various techniques in terms of pros and cons. Results are presented in terms of accuracy of detection for each technique. The results are discussed according to the recent problems and challenges in this field. The paper also highlights the applicability and reliability of the different techniques. Furthermore, some suggestions are presented for the future work in the field of driver drowsiness detection (DDD).

The security of healthcare and telemedicine systems is a critical issue that must be significantly investigated. Several smart telemedicine applications are expected to be adopted in the medical sector in the incoming years. Healthcare smart products that are connected through Internet to be accessible anytime and anywhere are expected to deal with critical and confidential information such as personal medical images. Therefore, medical image encryption is an important task in telemedicine and healthcare applications. This paper presents an efficient cryptosystem for medical image security based on exploiting the advantages of the de-oxyribo nucleic acid (DNA) rules and chaos maps. In the proposed medical image cryptosystem, logistic chaos map, piecewise linear chaotic map (PWLCM), and DNA encoding are employed. The PWLCM is employed to generate a secret key image. Then, the DNA rules are utilized for encoding the secret key image and the input plain image by rows that are encoded with the logistic chaos map. After that, the proposed logistic map is employed to obtain an intermediate image as another secret key image to set DNA functions row-by-row on the coded original image. Moreover, the intermediate image is decoded in the following stage. Finally, the previous actions are iterated through image columns once again to obtain the best ciphered image. The experimental results reveal that the suggested cryptosystem has a high security with an acceptable processing time. In addition, it can resist various kinds of attacks, such as known-plaintext and chosen-plaintext attacks.

The development of innovative methods for concealing critical data in multimedia files has exploded in information security in recent years. Cryptography and steganography cannot be used alone to protect data; rather, they can be combined and used in a single system. Audio steganography is among the most important information security techniques. It involves the concealment of information within audio signals to achieve covert communication. This paper introduces a comprehensive technique that integrates chaos Henon, Baker, and Arnold maps for image encryption with audio steganography to create a robust and secure audio steganography technique. First, the target image is encrypted using chaotic maps. Then, it is embeded within the high frequencies of the cover audio signal based on the Inverse Short Time Fourier Transform (ISTFT) to be transmitted to the destination through the channel. By integrating both encryption and concealment techniques, the cover audio signal quality can be preserved. Moreover, the hidden image security and robustness are improved, making the technique resistant to many types of attacks. The simulation results confirm that the suggested technique is robust in the presence of attacks. It achieves a distinct perceptual quality with an appreciated peak signal-to-noise ratio (PSNR) of 91.2 dB and a Mean Square Error (MSE) of 7.5 × 10 –10 . The randomness of the resulting encrypted image has successfully passed the National Institute of Standards and Technology (NIST) statistical test suite.

Visual information is extremely important in today’s world. Visual information transmitted in the form of digital images has become a critical mode of communication. As a result, digital image processing plays a critical role in advancing the image-related applications. Especially, in the medical field, the image processing stage is one of the important stages that need great accuracy to diagnose and determine the type of the disease. Its objective is to overcome the noise problems in medical images and preserve information and edges in images. Medical images can be enhanced by removing noise through the use of traditional and Deep Learning (DL) methods. DL methods depending on Convolutional Neural Networks (CNNs) have achieved great results in the processing stage for noise reduction in medical images. The DL is a promising and effective solution for estimating real noise and extracting representative features from images. This paper presents a review of image denoising methods for medical images, considering noise sources, and types of noise. The concepts of noise reduction (denoising) for various methods are presented. In addition, a comparative study is presented to clarify the advantages and disadvantages of each method. Finally, some possible trends for future work are introduced.

Effective medical diagnosis is dramatically expensive, especially in third-world countries. One of the common diseases is pneumonia, and because of the remarkable similarity between its types and the limited number of medical images for recent diseases related to pneumonia, the medical diagnosis of these diseases is a significant challenge. Hence, transfer learning represents a promising solution in transferring knowledge from generic tasks to specific tasks. Unfortunately, experimentation and utilization of different models of transfer learning do not achieve satisfactory results. In this study, we suggest the implementation of an automatic detection model, namely CADTra, to efficiently diagnose pneumonia-related diseases. This model is based on classification, denoising autoencoder, and transfer learning. Firstly, pre-processing is employed to prepare the medical images. It depends on an autoencoder denoising (AD) algorithm with a modified loss function depending on a Gaussian distribution for decoder output to maximize the chances for recovering inputs and clearly demonstrate their features, in order to improve the diagnosis process. Then, classification is performed using a transfer learning model and a four-layer convolution neural network (FCNN) to detect pneumonia. The proposed model supports binary classification of chest computed tomography (CT) images and multi-class classification of chest X-ray images. Finally, a comparative study is introduced for the classification performance with and without the denoising process. The proposed model achieves precisions of 98% and 99% for binary classification and multi-class classification, respectively, with the different ratios for training and testing. To demonstrate the efficiency and superiority of the proposed CADTra model, it is compared with some recent state-of-the-art CNN models. The achieved outcomes prove that the suggested model can help radiologists to detect pneumonia-related diseases and improve the diagnostic efficiency compared to the existing diagnosis models.

Nerve agents such as Sarin, Soman, and Tabun are among the most lethal chemical warfare agents, classified as mass destruction agents due to their extreme toxicity and rapid disruption of the nervous system. These highly volatile and easily dispersible compounds can be deployed in warfare or acts of terrorism, causing fatal respiratory failure, seizures, and irreversible nerve damage even at minimal exposure. The urgency of detecting these agents with high precision is critical for global security and counterterrorism efforts. To address this challenge, a highly sensitive photonic crystal fiber (PCF) sensor with an elliptical cladding and circular core (E-PCF) is designed for the rapid and accurate detection of nerve agents in the terahertz (THz) spectrum. The sensor employs circular air holes in the vestibule region to enhance light-matter interaction, optimizing detection through key performance metrics such as relative sensitivity, effective material loss, and confinement loss. Using two materials, such as silica glass and Zeonex as background materials, the proposed sensor demonstrates exceptional sensitivity and minimal loss. Numerical analysis within the 1.6–3.6 THz range reveals outstanding performance for Sarin (99.6% relative sensitivity, 3 × 10⁻ 13 dB/m confinement loss), Soman (98.8% relative sensitivity, 1.1 × 10⁻¹² dB/m loss), and Tabun (98% relative sensitivity, 7.6 × 10⁻ 11 dB/m loss). With its exceptional optical properties, silica glass ensures highly reliable detection, making the proposed sensor a powerful tool for counterterrorism efforts, environmental monitoring, industrial hazard detection, and military defense. This innovative PCF-based sensing technology marks a major breakthrough in chemical warfare agent detection, providing a fast, precise, and efficient solution for identifying highly toxic substances that pose severe threats to public safety and national security.

Crop productivity is crucial for farmers and economy worldwide. Factors such as fertilization, weather, and climate have a significant impact on yields. To improve crop productivity, a crop recommendation system is introduced in this paper. It provides data-driven advice on the best crops to plant, taking into account climate, weather, and soil nutrients. This research work introduces feature fusion with a stacking ensemble model comprising 18 classifiers and three novel methods to enhance crop recommendation and mitigate overfitting compared to other ensemble techniques. In this paper, we also examine two datasets for model validation; one of them is a large dataset containing nearly 28,242 records. The findings of our study reveal that feature fusion enables all ensemble classifiers to not only exceed the accuracy and precision of other established modern techniques, but also reduce overfitting, especially for the three proposed models that depend on a large dataset. In our experiments, the accuracy of ensemble models in categorizing diverse crops under different conditions ranges from 98.4% to 99.54%. Notably, the voting ensemble classifier proved to be the most effective, when applied to the first small dataset, achieving an impressive accuracy up to 99.56%. The second stacking ensemble classifier proved to be the most effective, when applied to the second large dataset, achieving an accuracy up to 85.6%.

Deep learning model training is known to be difficult and time-consuming. The principal component analysis network (PCANET) is a reasonably straightforward deep learning model that we use in this study to extract medical image features from multi-focus medical images (MFMI). The medical input images are divided into image features using a PCA filter, and activity level maps are constructed using the nuclear norm. This work applies nuclear norm and PCANET to build a usable feature space for medical image fusion. Particularly, the extracted PCANET characteristics can perform similarly to a convolutional neural network (CNN). Eventually, the final decision map (FDM) is assessed using the focus score map (FSM) produced by post-processing activities. The fused image is created by merging the two input medical images together using FDM. The PCANET features that are extracted in particular, can perform like a CNN-based network. According to the experimental outcomes, the findings of the fusion can better preserve the crucial data in the source images. This study shows the appearance of PCANET for a superior performance regarding the medical image fusion and good restoration quality. It also accomplishes the superior average values: AG of 8.419538, EI of 86.33734, and MI of 3.112971 for the SET1 (MRI and CT), in addition to the Q AB / F superior average values of 0.68963 for the SET 2 (MRI and SPECT). Finally, the Q CB superior average values of 0.754868, Q CV of 157.9377, and SF of 26.02398 are accomplished for the SET 3 (MRI and PET).