audio-diffusion / src /audio_to_images.py
teticio's picture
first commit
9c0c5c8
raw
history blame
1.45 kB
import os
import re
import json
import argparse
from tqdm.auto import tqdm
from mel import Mel
def main(args):
mel = Mel(x_res=args.resolution, y_res=args.resolution)
os.makedirs(args.output_dir, exist_ok=True)
audio_files = [
os.path.join(root, file)
for root, _, files in os.walk(args.input_dir)
for file in files
if re.search("\.(mp3|wav|m4a)$", file, re.IGNORECASE)
]
meta_data = {}
try:
for audio, audio_file in enumerate(tqdm(audio_files)):
try:
mel.load_audio(audio_file)
except KeyboardInterrupt:
raise
except:
continue
for slice in range(mel.get_number_of_slices()):
image = mel.audio_slice_to_image(slice)
image_file = f"{audio}_{slice}.png"
image.save(os.path.join(args.output_dir, image_file))
meta_data[image_file] = audio_file
finally:
with open(os.path.join(args.output_dir, 'meta_data.json'), 'wt') as file:
file.write(json.dumps(meta_data))
if __name__ == "__main__":
parser = argparse.ArgumentParser(description="Convert audio into Mel spectrograms.")
parser.add_argument("--input_dir", type=str)
parser.add_argument("--output_dir", type=str, default="data")
parser.add_argument("--resolution", type=int, default=256)
args = parser.parse_args()
main(args)