Open PhenomenalOnee opened 3 years ago
I am also interested in some of the other models that there are results for the paper. Especially the ones with lower time and frequency resolution (down to 31.25ms, 32 bands), as I am researching compression techniques for spectrograms.
Really Amazing stuff there Can you provide other pretrained models too like mobilnets for audio tagging Thank You