Automatic Video Dubbing

In recent years, the use of machine learning methods for task automation has become increasingly popular, including in the field of media content. This project focuses on developing a software solution for the automatic dubbing of videos from Russian to English to enhance the competitiveness of media platforms in the international market. The project addresses the key stages of automatic dubbing: speech-to-text conversion, text translation, speech synthesis, and audio-video synchronization. Modern neural network-based models were used to implement these tasks, ensuring high-quality data processing at each stage. The results demonstrate that the developed system provides high-quality dubbing, allowing foreign viewers to easily understand the translated content. Future improvements could include refining the audio synchronization method and enhancing the realism of the generated speech. This work confirms that it is possible to create a platform for high-quality automatic dubbing using existing methods, which can significantly expand the audience for video content platforms.