[FEATURE REQUEST] AudioSR: Versatile Audio Super-resolution at Scale

[ke1ne]

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

[gitmylo]

This could really help improve outputs from models such as AudioLDM, I'll add it to the backlog.

[madwurmz]

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.. 🔉