Music Cpasとは:
https://huggingface.co/datasets/google/MusicCaps
GrayScaleじゃないほうもあるから見てね(⋈◍>◡<◍)。✧♡(これ)
sampling_rate: int = 44100
20秒のwavファイル -> 1600×800のpngファイルへ変換
librosaの規格により、画像の縦軸:(0-10000?Hz), 画像の横軸:(0-40秒)
詳しくはlibrosa.specshow() ->
https://librosa.org/doc/main/auto_examples/plot_display.html
from datasets import load_dataset
data = load_dataset("mickylan2367/spectrogram")
data… See the full description on the dataset page:
https://huggingface.co/datasets/mickylan2367/GraySpectrogram.