Open ke1ne opened 9 months ago
This could really help improve outputs from models such as AudioLDM, I'll add it to the backlog.
This could really help improve outputs from models such as AudioLDM, I'll add it to the backlog.
this would be very very cool! I got some results and I can say it's very useful on 33k hz samples . Jonathan Fly was also looking into that repo! it needs some bug fixing and optimization.. 🔉
A new thing helping to "restore" low-res audio in a similar way as images https://audioldm.github.io/audiosr/ Has code too https://github.com/haoheliu/versatile_audio_super_resolution