Pose-based approaches for sign language recognition provide light-weight and fast models that can be adopted in real-time applications. This paper presented a framework for isolated Arabic sign language (ArSL) recognition using hand and face keypoints. We employed MediaPipe pose estimator for extracting the keypoints of sign gestures in the video stream. Using the extracted keypoints, three models were proposed for sign language recognition, Long-Term Short Memory (LSTM), Temporal Convolution Networks (TCN) and Transformer based models. Moreover, we investigated the importance of non-manual features for sign language recognition systems and the obtained results showed that combining hand and face keypoints boosted the recognition accuracy by around \(4\% \) compared with only hand keypoints. The proposed models were evaluated on Arabic and Argentinian sign languages. Using the KArSL-100 dataset, the proposed pose-based Transformer achieved the highest accuracy of \(99.74\% \) and \(68.2\% \) in signer-dependent and independent modes, respectively. Additionally, the Transformer was evaluated on the LSA64 dataset and obtained an accuracy of \(98.25\% \) and \(91.09\% \) in signer-dependent and independent modes, respectively. Consequently, the pose-based Transformer outperformed the state-of-the-art techniques on both datasets using keypoints from the signer’s hands and face.

Sentiment classification is the process of classifying emotions and opinions in texts. In this study, the problem of Arabic sentiment analysis was addressed. A support vector machine (SVM) model was proposed to classify opinions in Arabic micro-texts as being positive or negative. To evaluate the performance of the SVM model, a dataset was built from tweets discussing several social issues in Saudi Arabia. These issues include changes that were implemented by the country as part of a newly established vision, known as Saudi Arabia Vision 2030. The constructed dataset was manually annotated according to the sentiment conveyed in the text. To achieve the best sentiment classification accuracy, several procedures were implemented within the proposed framework including light stemming, feature extraction (Ngrams, emoji and tweet-topic features), parameter optimisation and feature-set reduction. The experimental results revealed excellent outcomes. An accuracy of 89.83% was achieved using the proposed SVM model.

No abstract

The Internet of things is a popular concept in the current digital revolution. Nowadays, devices worldwide can be connected to the Internet, enhancing their communication, capabilities, and intelligence. Low-Power Wireless Personal Area Network (6LoWPAN) was specifically designed to build wireless networks for IoT resource-constrained devices. However, 6LoWPAN is susceptible to several security attacks. The fragmentation mechanism, in particular, is vulnerable to various attacks due to the lack of fragment authentication and verification procedures in the adaptation layer. This article provides a survey of fragmentation attacks and available countermeasures. Furthermore, the buffer reservation attack, one of the most harmful fragmentation attacks that may cause DoS, is studied and simulated in detail. A countermeasure for this attack is also implemented based on a reputation-scoring scheme. Experiments showed the harmful effects of the buffer reservation attack and the effectiveness of the implemented reputation-scoring countermeasure.