Sign up

drbaph/NovaSR

🤗 On Hugging Faceaudio-to-audioapache-2.01 MBother✓ HF checksums availableupdated today
Magnet

Custom Node for ComfyUI

https://github.com/Saganaki22/ComfyUI-NovaSR

!image

NovaSR: Pushing the Limits of Extreme Efficiency in Audio Super-Resolution

This is the model for NovaSR, a tiny 50kb audio upsampling model that upscales muffled 16khz audio into clear and crisp 48khz audio at speeds from 100-3500x realtime.

Audio Samples

Before Processing (16kHz):

Your browser does not support the audio element.

After Processing (48kHz):

Your browser does not support the audio element.

!ComfyUI_temp_tepqs_00001_

Details

  • Model Size: 52kb for pytorch version
  • Input Rate: 16kHz
  • Output Rate: 48kHz
  • Inference Speed: 300-3500x realtime depending on gpu
  • Mono

Comparisons

Comparisons were done on A100 gpu. Higher realtime means faster processing speeds.

Comparison on CPU are coming soon.

| Model | Speed (Real-Time) | Model Size |

| :------------ | :---------------- | :--------- |

| NovaSR | 3600x realtime | ~52 KB |

| FlowHigh | 20x realtime | ~450 MB |

| FlashSR | 14x realtime | ~1000 MB |

| AudioSR | 0.6x realtime | ~6000 MB |

Usage

Please check out the github repo for usage: https://github.com/Saganaki22/ComfyUI-NovaSR

Original Repo: https://github.com/ysharma3501/NovaSR

If you find the model/code helpful, stars or likes would be appreciated.

Thank you.