Mohammad Sarosh Umar’s research while affiliated with Aligarh Muslim University and other places

The rapid growth of biometric authentication systems has heightened the need to address privacy and security concerns. The cancellable biometric template generation is one of the promising solutions in this situation which protects both the system and the user’s biometric data from unauthorized access ensuring non-invertibility, unlinkability, and secure template encryption. The random projection-based cancellable biometric template generation is one of the efficient techniques to secure the information of users. However, random projection-based approaches remain vulnerable to attacks where repeated projections may reconstruct the original template. This paper introduces MCC Encoded Random Triangle Hashing, which encodes the projection matrix using minutiae cylindrical codes to prevent template reconstruction in case of matrix leakage. The proposed method is evaluated on six publicly available fingerprint databases FVC2000 DB1, FVC2000 DB2, FVC2002 DB1, FVC2002 DB3, FVC2004 DB1, and FVC2004 DB3 using FMR, FAR, GAR, ROC, and EER metrics, demonstrating improved template security.

Melanoma is a highly malignant skin cancer that may be fatal if not promptly detected and treated. The limited availability of high-quality melanoma images, which are needed for training machine learning models, is one of the obstacles to detecting melanoma. Generative adversarial networks (GANs) have grown in popularity as a strong technique for image synthesis. This research is also targeted at the sustainable development goal (SDG) for health care. In this study, we survey existing GAN-based melanoma image synthesis methods. In this work, we briefly introduce GANs and how they may be used for generating synthetic images. Ensuring healthy lifestyles and promoting well-being for everyone, regardless of age, is the main aim. A comparative study is carried out on how GANs are used in current research to generate melanoma images and how they improve the classification performance of neural networks. Various public and proprietary datasets for training GANs in melanoma image synthesis are also discussed. Lastly, we assess the examined studies' performance using measures like the Frechet Inception distance (FID), Inception score, structural similarity ındex (SSIM), and various classification performance metrics. We compare the evaluated findings and suggest further GAN-based melanoma image-creation research.

Mobile devices are currently more susceptible to cyberattacks than other types of devices, such as personal computers, as a result of the effects and development and their common usage in daily life. Mobile phones have text messaging capabilities, often known as SMS (Short Text Messages), which are used by cybercriminals to target people. A malicious program that uses text messages to target smartphone users is frequently referred to as SMS phishing. Smishing is a sort of phishing, although there are several ways in which it differs from phishing, including the amount of information that is included in the SMS and the offensive approach. Then 50 literature works are examined that concentrate on smishing detection and prevention related to the mobile application for this survey to highlight the numerous shortcomings of the present methodologies for smishing prevention and identification in mobile apps. The study of numerous approaches will be clarified in this instance based on many aspects, like significant metrics, publication year and journals, numerical values, and other criteria. This paper also discusses possibilities for future research as well as the difficulties encountered with the detection techniques currently employed in mobile applications.

Phishing detection technologies are essential to ensuring that users have a secure online experience by preventing users from falling prey to online fraud, divulging personal information to an attacker, and other risks. To stop e-mail scams, this research employs the FFO-LSTM classifier to identify phishing websites. The FFO method is used to pick and optimise the relevant parameters from many parameters that the LSTM offers. This reduces the computational complexity of the system and improves performance. The LSTM classifier conducts a more thorough examination of the network, which aids in enhancing the effectiveness of phishing detection. The FFO-LSTM classifier achieved exceptional values of 93.16%, 94.59%, and 92.66%, which illustrates the efficiency of the research. The performance metrics are utilised to illustrate the significance of the research.

Phishing is a malicious technique employed by hackers to illicitly obtain confidential user information for their own misuse. In contemporary times, three notable approaches have emerged as effective countermeasures against such attacks: awareness-focused strategies, blacklist-based methods, and machine learning (ML) techniques. Hence, this systematic paper aims to provide researchers, readers, and users with an analysis of diverse proposals from other scholars on how to effectively combat phishing using ML algorithms. The research focuses on phishing detection utilizing ML techniques, establishing a taxonomy of machine learning models by examining their respective advantages and disadvantages. Furthermore, the performance analysis based on the available datasets of the ML classifiers such as NB, SVM, KNN, RF, and ANN is evaluated in terms of the performance metrics and the accuracy values of the classifiers on kfold at 10 is 82.40%, 83.00%, 83.60%, 84.39%, and 84.84%, and TP at 8 is 81.03%, 81.62%, 82.21%, 83.23%, and 83.77%, respectively.