Nowadays, cyber attackers focus on Android, which is the most popular open-source operating system, as main target by applying some malicious software (malware) to access users' private information, control the device, or harm end-users. To detect An...

Gerçek zamanlı çalışan sistemlerde görüntü işleme uygulamaları yapmak son zamanlarda oldukça popüler olan bir konu haline gelmiştir. Yapay zekâ alanının alt dallarından biri olan derin öğrenme yöntemleri ve görüntülerden nesne tespiti yapma alanında kullanılan görüntü işleme algoritmaları birlikte kullanılarak, otonom otomobiller, otonom insansız hava araçları, yardımcı robot teknolojileri, engelli ve yaşlı bireyler için asistan teknolojileri gibi birçok alanda uygulamalar geliştirilmektedir. Yapılan çalışmada, görme engelli bireyler, otonom araçlar ve robotlar tarafından kullanılabilecek yardımcı bir teknoloji sistemi tasarlamak için dokunsal parke yüzeylerinin derin öğrenme yöntemleriyle tespit edilmesi gerçekleştirilmiştir. Geleneksel görüntü işleme algoritmalarının aksine bu çalışmada derin öğrenme yöntemleri ile görüntü işleme algoritmaları birlikte kullanılmıştır. Nesne tespit etme yöntemleri içinde en iyi yöntemlerden biri olan You Only Look Once-V3(YOLO-V3) modeli DenseNet modeli ile birleştirilerek YOLOV3-Dense modeli oluşturulmuştur. YOLO-V2, YOLO-V3 ve YOLOV3Dense modelleri tarafımızca oluşturulmuş olan ve içerisinde 4580 etiketli görsel bulunan Marmara Dokunsal Parke Yüzeyi(MDPY) veri seti üzerinde ayrı ayrı eğitildikten sonra performansları test veri seti üzerinde birbirleri ile karşılaştırılmıştır. %89 F1-skor, %92 ortalama hassasiyet ve %81 IoU değerleri ile YOLOV3-Dense modelinin dokunsal parke yüzeyi tespit etmede diğer modellerden daha iyi olduğu gözlemlenmiştir. Saniyede 60 kare çalışma hızı ile YOLOV3-Dense modeli gerçek zamanlı çalışan sistemlerde de kullanılabilmektedir.

Fabrication and characterization of flexible optical fiber bundles (FBs) with in-house synthesized high-index and low-index thermally matched glasses are presented. The FBs composed of around 15000 single-core fibers with pixel sizes between 1.1 and 10 μm are fabricated using the stack-and-draw technique from sets of thermally matched zirconium-silicate ZR3, borosilicate SK222, sodium-silicate K209, and F2 glasses. With high refractive index contrast pair of glasses ZR3/SK222 and K209/F2, FBs with numerical apertures (NAs) of 0.53 and 0.59 are obtained, respectively. Among the studied glass materials, ZR3, SK222, and K209 are in-house synthesized, while F2 is commercially acquired. Seven different FBs with varying pixel sizes and bundle diameters are characterized. Brightfield imaging of a micro-ruler and a Convallaria majalis sample and fluorescence imaging of a dye-stained paper tissue and a cirrhotic mice liver tissue are demonstrated using these FBs, demonstrating their good potential for microendoscopic imaging. Brightfield and fluorescence imaging performance of the studied FBs are compared. For both sets of glass compositions, good imaging performance is observed for FBs, with core diameter and core-to-core distance values larger than 1.6 μm and 2.3 μm, respectively. FBs fabricated with K209/F2 glass pairs revealed better performance in fluorescence imaging due to their higher NA of 0.59.

In today’s world, accurately understanding and interpreting emotions in human-computer interaction is important. In this context, this study has adopted a detailed approach to the emotion recognition problem on both speech and text data using the Interactive Emotional Dyadic Motion Capture (IEMOCAP) dataset. First, the problem of datasets with limited number of records and unbalanced distribution across classes was addressed. For this purpose, a dataset obtained from records created as improvised in the IEMOCAP dataset was used and data augmentation methods were applied for both speech and text data. Using datasets that were balanced by applying data augmentation, single-mode emotion recognition experiments were performed with models developed for Speech Emotion Recognition (SER) and Textual Emotion Recognition (TER). Subsequently, the features obtained from these two single modalities were combined with the intermediate fusion method to provide more comprehensive emotion recognition and accuracy, and the Bimodal Emotion Recognition (BiMER) system was developed. The ResNet50-CRNN+AT model, which we obtained the highest accuracy from the three different models developed for SER, creates the speech mode of BiMER, while the Bidirectional Encoder Representations from Transformers (BERT) model used for TER creates the text mode of BiMER. In this way, BiMER was supported with data augmentation methods and the robustness and generalization ability of the model were improved, reaching 88.33% accuracy. Finally, the developed BiMER system was implemented as a real-time web application using the Flask framework, and the capacity of this application to recognize emotions interactively through the user interface was tested.

Advancements in microelectronics have increased the popularity of mobile devices like cellphones, tablets, e-readers, and PDAs. Android, with its open-source platform, broad device support, customizability, and integration with the Google ecosystem, has become the leading operating system for mobile devices. While Android’s openness brings benefits, it has downsides like a lack of official support, fragmentation, complexity, and security risks if not maintained. Malware exploits these vulnerabilities for unauthorized actions and data theft. To enhance device security, static and dynamic analysis techniques can be employed. However, current attackers are becoming increasingly sophisticated, and they are employing packaging, code obfuscation, and encryption techniques to evade detection models. Researchers prefer flexible artificial intelligence methods, particularly deep learning models, for detecting and classifying malware on Android systems. In this survey study, a detailed literature review was conducted to investigate and analyze how deep learning approaches have been applied to malware detection on Android systems. The study also provides an overview of the Android architecture, datasets used for deep learning-based detection, and open issues that will be studied in the future.