Hugging Face's logo

tnkmr
/
sfi_convtasnet_fd_mgf_musdb18hq

Audio-to-Audio
Japanese
music
audio
SFI
Model card Files Community
4
tnkmr commited on
Commit
7eb2370
•
1 Parent(s): 1dbfab8

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -0
README.md CHANGED
@@ -15,6 +15,7 @@ metrics:
15
 
16
  # Sampling-frequency-independent (SFI) Conv-TasNet trained with the MUSDB18-HQ dataset for music source separation.
17
  This model was proposed in [our IEEE/ACM Trans. ASLP paper](https://doi.org/10.1109/TASLP.2022.3203907) and works well with untrained sampling frequencies by using sampling-frequency-independent convolutional layers with the frequency domain filter design.
 
18
  It was trained by Tomohiko Nakamura using [the codebase](https://github.com/TomohikoNakamura/sfi_convtasnet)).
19
  This model was trained with 32 kHz-sampled data but works well with untrained sampling frequencies (e.g., 8, 16 kHz).
20
 
 
15
 
16
  # Sampling-frequency-independent (SFI) Conv-TasNet trained with the MUSDB18-HQ dataset for music source separation.
17
  This model was proposed in [our IEEE/ACM Trans. ASLP paper](https://doi.org/10.1109/TASLP.2022.3203907) and works well with untrained sampling frequencies by using sampling-frequency-independent convolutional layers with the frequency domain filter design.
18
+ The latent analog filter is a modulated Gaussian filter.
19
  It was trained by Tomohiko Nakamura using [the codebase](https://github.com/TomohikoNakamura/sfi_convtasnet)).
20
  This model was trained with 32 kHz-sampled data but works well with untrained sampling frequencies (e.g., 8, 16 kHz).
21