New project
Added 2024-12-27 21:49:02 +0000 UTCI also started writing my own translator. This will be necessary if, for example, you do not speak English (but you know, for example, Russian or Polish, etc.). What it does: downloads a video from YouTube (for example, a tutorial in Spanish), then extracts the audio channel, the first neural network recognizes the words and writes them to a file for subtitles, the second neural network translates into the desired language and performs voice cloning and vocalizes the text, and tries to clone the intonation, and at the end of the video the generated The voice. At the end, the program replaces the original audio track with another one. I think this project will be useful to everyone, and it will be on GitHub and open source. I also want to add that the neural network data takes up 1 gigabyte of video memory (I checked it myself), but it's still inaccurate data, since I ran them separately.
I'm waiting for your opinion, gentlemen
:3
