Real Time Video Changer

• 

  First versions of many to come... Im running out of time... MORE TIME

Inspiration

Thousands of videos with great content, but the audio, Help!!!!

What it does

Get a video and change the audio in real time, just with a better speech accent.

How we built it

1. First, choose an example video, with that what we want to improve: an English rare accent.
2. Get the transcription of that video, and save it to a file.
3. Now, that transcription turn into a *.wav or any audio file, to put in the original video to replace the audio with the one just improved.
4. Done, now, view the entire video, with our improved audio in English Language.

Challenges we ran into

1. Get a suitable AI Model, free of charge, and with low latency.
2. Will be needed 2 (two) models: one for transcript the original audio, then other for turn that transcription into audio and overwrite the original.
3. better work over video directly, extract original audio, improve it, and then save a new video with the new audio.

Accomplishments that we're proud of

Not yet, I hope soon.

What we learned

What's next for Real Time Video Audio Improvement

Built With

• docker
• docker-desktop
• nvidia
• nvidia-workbench
• python
• pytorch
• real-time
• voice-accent
• voice-audio
• workbench