This paper presents an AI-based pipeline that translates English YouTube videos into Uzbek and produces synchronized dubbed speech in near real time. The system integrates audio acquisition, preprocessing (16 kHz mono PCM), automatic speech recognition with OpenAI Whisper, neural machine translation, and text-to-speech synthesis. A task-queue architecture with status and log monitoring improves robustness for long videos, while fallback translation and TTS services maintain continuity under failures. Experiments on diverse video genres demonstrate practical latency and intelligible Uzbek output, making the approach suitable for localization, education, and media applications.
| Mualliflar | Ахмедов, Муроджон, Омаджон, Уришев |
|---|---|
| Jurnal | Al-Farg'oniy avlodlari |
| Nashr sanasi | 2026-02-28 |
| Son | 1 |
| Betlar | 74-79 |
| Til | Rus |
искусственный интеллект, распознавание речи, машинный перевод, синтез речи, перевод в реальном времени, мультимедийный контент, artificial intelligence, speech recognition, machine translation, speech synthesis, , real-time translation, multimedia content
The article proposes complex mathematical models for probabilistic assessment of learning process indicators, the distinguishing feature of which is the use of the following learning process indicators: the probability…
Mazkur maqolada yorug‘lik nurlovchi diodlarni ishlab chiqarishda qo‘llaniladigan asosiy texnologiyalar tahlil qilinadi. AIII–VV turdagi yarimo‘tkazgich birikmalari asosida epitaksial qatlamlar o‘stirishning gaz fazali…
This article presents a review of existing battery technologies used in battery energy storage systems, with particular emphasis on their technical characteristics, operational limitations, and suitability for…
The objective of this study is to evaluate the performance of WebSocket as an intermediate protocol for integrating digital twins with Siemens S7 PLCs (Programmable Logic Controllers) in remote environments and compare…
Finding optimal offloading decisions for dependent tasks with multiple objectives and hard deadlines in resource-constrained mobile edge computing networks is a challenging problem. The solution to this problem is…
The method presented in this article has potential applications in computer science and can be used to create a new algorithm for factoring numbers or to further improve the efficiency of existing algorithms. To do…
In this paper, the computational efficiency of parametric exponentiation operations within the framework of parametric algebra is analyzed, and an algorithmic approach for their fast computation is proposed. By…
This article presents a comparative analysis of the effectiveness of cryptographic and steganographic methods in the process of covert information transmission in a digital environment. During the study, the operating…
Early detection and effective extinguishing of fires are important factors in ensuring safety in closed environments. In this article, algorithms for an intelligent system based on camera systems are developed to detect…
This article describes a new method for dividing integers into their prime factors. The oldest and most classical method for determining whether a number is prime or not is to divide it into primes in sequence, that is…