Spaces:
Running on Zero
Running on Zero
File size: 3,555 Bytes
127b976 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 |
from __future__ import annotations
import argparse
import random
from pathlib import Path
import pandas as pd
AUDIO_EXTENSIONS = {
".wav", ".flac", ".mp3", ".m4a", ".ogg", ".opus", ".aac"
}
def find_audio_files(audio_root: Path):
files = []
for p in audio_root.rglob("*"):
if p.is_file() and p.suffix.lower() in AUDIO_EXTENSIONS:
files.append(p)
return sorted(files)
def make_rows(files, split_name: str):
rows = []
for p in files:
rows.append({
"file_path": p.resolve().as_posix(),
"binary_label": 0,
"attack_type": "bonafide",
"start_fake": -1,
"end_fake": -1,
"dataset": "GLOBE",
"split": split_name,
"source_file": p.name,
"source_folder": p.parent.name,
})
return rows
def save_csv(rows, out_csv: Path):
out_csv.parent.mkdir(parents=True, exist_ok=True)
df = pd.DataFrame(rows)
df.to_csv(out_csv, index=False)
print(f"Saved: {out_csv}")
print(f"Rows: {len(df)}")
if len(df):
print(df[["dataset", "split", "binary_label", "attack_type"]].value_counts())
def main():
parser = argparse.ArgumentParser(
description=(
"Create small real/bonafide GLOBE metadata CSVs for AuralGuard. "
"This does not train anything; it only creates train/val/test CSV files."
)
)
parser.add_argument(
"--audio-root",
required=True,
help="Folder containing GLOBE audio files. The script searches recursively."
)
parser.add_argument(
"--out-dir",
default="data/metadata",
help="Where to save train_globe.csv, val_globe.csv, globe_test.csv"
)
parser.add_argument("--max-train", type=int, default=2000)
parser.add_argument("--max-val", type=int, default=300)
parser.add_argument("--max-test", type=int, default=300)
parser.add_argument("--seed", type=int, default=42)
args = parser.parse_args()
audio_root = Path(args.audio_root)
if not audio_root.exists():
raise FileNotFoundError(f"Audio root does not exist: {audio_root}")
files = find_audio_files(audio_root)
print(f"Found audio files: {len(files)}")
if len(files) == 0:
raise RuntimeError(
"No audio files found. Check the folder path or supported extensions: "
+ ", ".join(sorted(AUDIO_EXTENSIONS))
)
rng = random.Random(args.seed)
rng.shuffle(files)
n_train = min(args.max_train, len(files))
n_val = min(args.max_val, max(0, len(files) - n_train))
n_test = min(args.max_test, max(0, len(files) - n_train - n_val))
train_files = files[:n_train]
val_files = files[n_train:n_train + n_val]
test_files = files[n_train + n_val:n_train + n_val + n_test]
out_dir = Path(args.out_dir)
save_csv(make_rows(train_files, "train"), out_dir / "train_globe.csv")
save_csv(make_rows(val_files, "val"), out_dir / "val_globe.csv")
save_csv(make_rows(test_files, "test"), out_dir / "globe_test.csv")
print("\nDone.")
print("Recommended next step:")
print('python scripts\\check_metadata.py --csv "data\\metadata\\train_globe.csv"')
print('python scripts\\check_metadata.py --csv "data\\metadata\\val_globe.csv"')
print('python scripts\\check_metadata.py --csv "data\\metadata\\globe_test.csv"')
if __name__ == "__main__":
main()
|