This scientific article analyzes multimodal security mechanismsbased on voice and video synthesis. Due to the increasing importance ofbiometric authentication methods in modern information systems, the issue ofjoint processing of audio and video data is relevant. The research used modelsbased on artificial intelligence, in particular convolutional neural networks(CNN) and recurrent neural networks (LSTM). The accuracy and reliability of thesystem were increased through a multimodal approach. The risks of deep faketechnologies and mechanisms for protecting against them were also considered.The results show that multimodal systems have high efficiency and justify theirwidespread use in the security field.
| Mualliflar | G'oipova Kh |
|---|---|
| Jurnal | Management and Economics Scientific Research Journal |
| Nashr sanasi | 2026-05-21 |
| Jild | 3 |
| Son | 3 |
| Betlar | 17-22 |
| Til | Ingliz |
voice synthesis, video synthesis, biometric authentication, multimodal system, artificial intelligence, CNN, LSTM, deep fake, facial recognition, voice recognition, signal processing, information security, identification, fusion, neural networks
This paper explores advanced methods for generalizingamplitude amplification in quantum algorithms beyond the traditional Grover'sAlgorithm framework. The study demonstrates how generalized amplitudeamplification…
This in the article voice analysis in technologies applicablespecial algorithms and artificial intellect models is studied. Speech signals againperformance , acoustic features separate to take and mechanical study…
This article analyzes the didactic advantages of switching fromthe traditional AutoCAD program to the Revit program based on BIMtechnologies in teaching engineering graphics. During the study, the capabilitiesof AutoCAD…
Rural medical centers (RMCs, now family doctor stations),which are the main link in primary health care in rural areas of Uzbekistan, playan important role in providing medical services close to the population.However…
This article analyzes the possibilities of using artificialintelligence and deep learning algorithms for the automatic detection of diseasesbased on eye images. In the study, transfer learning technology based on the…
This article examines the impact of accelerated demographic growth on the stateand structure of the labor market in Uzbekistan during 2010–2023. Based on data from the StateStatistics Committee of Uzbekistan, the…
As everyone knows, the Internet is currently the most effectivemeans of communication, a means of reaching a large audience in a short periodof time (a very short period for history). This type of communication…
This article reviews information on the processes of applying artificial intelligencetechnologies in the mechanical engineering industry and their impact on production efficiency. It alsostudies foreign experience in…
This thesis discusses the problem of improving low-qualityimages using machine learning and computer vision technologies. The studyanalyzes the Real-ESRGAN algorithm based on generative adversarial networksand its…
In the context of intensifying climate change, resource scarcity, and wideningdisparities in regional development, improving the mechanism for supporting the use of greentechnologies to ensure sustainable economic…
Management and Economics Scientific Research Journal — barcha maqolalar