Sabah Abdulazeez Jebur

Detecting violence in various scenarios is a difficult task that requires a high degree of generalisation. This includes fights in different environments such as schools, streets, and football stadiums. However, most current research on violence detection focuses on a single scenario, limiting its ability to generalise across multiple scenarios. To tackle this issue, this paper offers a new multi-scenario violence detection framework that operates in two environments: fighting in various locations and rugby stadiums. This framework has three main steps. Firstly, it uses transfer learning by employing three pre-trained models from the ImageNet dataset: Xception, Inception, and InceptionResNet. This approach enhances generalisation and prevents overfitting, as these models have already learned valuable features from a large and diverse dataset. Secondly, the framework combines features extracted from the three models through feature fusion, which improves feature representation and enhances performance. Lastly, the concatenation step combines the features of the first violence scenario with the second scenario to train a machine learning classifier, enabling the classifier to generalise across both scenarios. This concatenation framework is highly flexible, as it can incorporate multiple violence scenarios without requiring training from scratch with additional scenarios. The Fusion model, which incorporates feature fusion from multiple models, obtained an accuracy of 97.66% on the RLVS dataset and 92.89% on the Hockey dataset. The Concatenation model accomplished an accuracy of 97.64% on the RLVS and 92.41% on the Hockey datasets with just a single classifier. This is the first framework that allows for the classification of multiple violent scenarios within a single classifier. Furthermore, this framework is not limited to violence detection and can be adapted to different tasks.

The use of video surveillance systems has increased due to security concerns and their relatively low cost. Researchers are working to create intelligent Closed Circuit Television (CCTV) cameras that can automatically analyze behavior in real-time to detect anomalous behaviors and prevent dangerous accidents. Deep Learning (DL) approaches, particularly Convolutional Neural Networks (CNNs), have shown outstanding results in video analysis and anomaly detection. This research paper focused on using Inception-v3 transfer learning approaches to improve the accuracy and efficiency of abnormal behavior detection in video surveillance. The Inception-v3 network is used to classify keyframes of a video as normal or abnormal behaviors by utilizing both pre-training and fine-tuning transfer learning approaches to extract features from the input data and develop a new classifier. The UCF-Crime dataset is used to train and evaluate the proposed models. The performance of both models was evaluated using accuracy, recall, precision, and F1 score. The fine-tuned model achieved 88.0%, 89.24%, 85.83%, and 87.50% for these measures, respectively. In contrast, the pre-trained model obtained 86.2%, 86.43%, 84.62%, and 85.52%, respectively. These results demonstrate that transfer learning using Inception-v3 architecture can effectively classify normal and abnormal behaviors in videos, and fine-tuning the weights of the layers can further improve the model's performance. Index Terms— Abnormal behavior detection, Video surveillance, Deep learning, Transfer learning, InceptionV3.

The highest way to protect data from intruder and unauthorized persons has become a major issue. This matter led to the development of many techniques for data security, such as Steganography, Cryptography, and Watermarking to disguise data. This paper proposes an image steganography method using the Least Significant Bits (LSB) technique and XOR operator and a secret key, through which the secret key is transformed into a one-dimensional bit stream array, then these bits are XORed with the bits of the secret image. Multiple experiments have been performed to embed color and grayscale images inside cover media. In this work, the LSB technique is ideal in two ways: firstly, only the least significant one-bit (1bit) of each byte will store the embedded data, this method is named (1-LSB). Secondly, the four least significant bits of the right half-byte (4 bits) of each byte will store the embedded data, this method is named (4-LSB). Subjective and objective analyzes were performed for each LSB process. The subjective analysis is responsible for both HVS and histogram, whereas the objective analysis involved both PSNR and MSE metrics.

In the last few years, due to the continuous advancement of technology, human behavior detection and recognition have become important scientific research in the field of computer vision (CV). However, one of the most challenging problems in CV is anomaly detection (AD) because of the complex environment and the difficulty in extracting a particular feature that correlates with a particular event. As the number of cameras monitoring a given area increases, it will become vital to have systems capable of learning from the vast amounts of available data to identify any potential suspicious behavior. Then, the introduction of deep learning (DL) has brought new development directions for AD. In particular, DL models such as convolution neural networks (CNNs) and recurrent neural networks (RNNs) have achieved excellent performance dealing with AD tasks, as well as other challenging domains like image classification, object detection, and speech processing. In this review, we aim to present a comprehensive overview of those research methods using DL to address the AD problem. Firstly, different classifications of anomalies are introduced, and then the DL methods and architectures used for video AD are discussed and analyzed, respectively. The revised contributions have been categorized by the network type, architecture model, datasets, and performance metrics that are used to evaluate these methodologies. Moreover, several applications of video AD have been discussed. Finally, we outlined the challenges and future directions for further research in the field.

Corona virus’s correct and accurate diagnosis is the most important reason for contributing to the treatment of this disease. Radiography is one of the simplest methods to detect virus infection. In this research, a method has been proposed that can diagnose disease based on radiography (X-ray chest) and deep learning techniques. We conducted a comparative study by using three diagnosis models; the first one was developed by using traditional CNN, while the two others are our proposed models (second and third models). The proposed models can diagnose the COVID-19 infection, normal cases, lung opacity, and Viral Pneumonia according to the four categories in the covid19 radiography dataset. The transfer learning technology had used to increase the robustness and reliability of our model, also, data augmentation was used for reducing the overfitting and to increase the accuracy of the model by scaling rotation, zooming, and translation. The third model showed higher training accuracy of 93.18% compared to the two other models that are dependent on using traditional convolution neural networks with an accuracy of 70.28% of the first model, while the accuracy of the second model that uses data augmentation with traditional convolution neural is 90.1%, while the testing accuracy models was 68.27% for the first model, 87.55% for the second model, and 86.03% for the third model.