Spaces:
Running on Zero
Running on Zero
File size: 1,518 Bytes
b8ced22 | 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 | """Pack aligned RGB frames and upstream teacher outputs into a validated clip."""
import argparse
from pathlib import Path
import sys
sys.path.insert(0, str(Path(__file__).resolve().parents[1]))
import numpy as np
from pure.images import load_images
from pure.training.cache import ARRAY_KEYS, write_cache
def main():
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument(
"--images",
nargs="+",
required=True,
help="Frame paths in temporal order; no implicit resizing",
)
parser.add_argument(
"--labels",
required=True,
help="NPZ containing the aligned depth/ray/mask/match arrays",
)
parser.add_argument("--output", required=True)
parser.add_argument(
"--teacher-prediction",
help="Optional infer.py NPZ from the frozen base; provides teacher_xy",
)
args = parser.parse_args()
images = load_images(args.images)
with np.load(args.labels, allow_pickle=False) as archive:
arrays = {
key: archive[key].copy()
for key in ARRAY_KEYS
if key != "images" and not (key == "teacher_xy" and args.teacher_prediction)
}
if args.teacher_prediction:
with np.load(args.teacher_prediction, allow_pickle=False) as pred:
arrays["teacher_xy"] = pred["xy"].copy()
arrays["images"] = images
write_cache(args.output, arrays)
print(f"Validated clip saved: {args.output}")
if __name__ == "__main__":
main()
|