File size: 4,321 Bytes
b0b8755 | 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 | #!/usr/bin/env python3
"""Validate the streaming-readable XS-VID v2 release annotation contract."""
from __future__ import annotations
import argparse
import hashlib
import json
import math
from collections import Counter
from pathlib import Path
from build_unified_annotations import JsonArrayReader
def sha256(path: Path) -> str:
digest = hashlib.sha256()
with path.open("rb") as handle:
for block in iter(lambda: handle.read(16 * 1024 * 1024), b""):
digest.update(block)
return digest.hexdigest()
def main() -> None:
parser = argparse.ArgumentParser()
parser.add_argument("annotation", type=Path)
parser.add_argument("--output", type=Path, required=True)
args = parser.parse_args()
counts: dict[str, int] = {}
ids: dict[str, set[int]] = {}
image_sizes: dict[int, tuple[int, int]] = {}
tracks: dict[int, dict] = {}
errors: list[str] = []
for collection in ("videos", "images", "annotations", "categories", "tracks"):
values: set[int] = set()
previous_id = -1
for record in JsonArrayReader(args.annotation, collection):
record_id = record.get("id")
if not isinstance(record_id, int):
errors.append(f"{collection} record without integer id")
continue
if record_id in values:
errors.append(f"duplicate {collection} id {record_id}")
if record_id <= previous_id:
errors.append(f"non-increasing {collection} id {record_id}")
values.add(record_id)
previous_id = record_id
if collection == "annotations":
if not isinstance(record.get("track_id"), int):
errors.append(f"annotation {record_id} lacks integer track_id")
elif collection == "tracks":
tracks[record_id] = record
elif collection == "images":
image_sizes[record_id] = (record.get("width", 0), record.get("height", 0))
counts[collection] = len(values)
ids[collection] = values
dangling = Counter()
for image in JsonArrayReader(args.annotation, "images"):
if image.get("video_id") not in ids["videos"]:
dangling["image_video"] += 1
for ann in JsonArrayReader(args.annotation, "annotations"):
if ann.get("image_id") not in ids["images"]:
dangling["annotation_image"] += 1
if ann.get("video_id") not in ids["videos"]:
dangling["annotation_video"] += 1
if ann.get("category_id") not in ids["categories"]:
dangling["annotation_category"] += 1
if ann.get("track_id") not in ids["tracks"]:
dangling["annotation_track"] += 1
else:
track = tracks[ann["track_id"]]
if track.get("video_id") != ann.get("video_id"):
dangling["annotation_track_video"] += 1
if track.get("category_id") != ann.get("category_id"):
dangling["annotation_track_category"] += 1
bbox = ann.get("bbox")
image_size = image_sizes.get(ann.get("image_id"))
if not isinstance(bbox, list) or len(bbox) != 4 or image_size is None:
dangling["annotation_bbox"] += 1
elif not all(isinstance(value, (int, float)) and math.isfinite(value) for value in bbox):
dangling["annotation_bbox"] += 1
else:
x, y, width, height = bbox
image_width, image_height = image_size
if width <= 0 or height <= 0 or x < 0 or y < 0 or x + width > image_width or y + height > image_height:
dangling["annotation_bbox"] += 1
for name, count in dangling.items():
if count:
errors.append(f"{count} dangling {name} references")
result = {
"annotation": args.annotation.name,
"sha256": sha256(args.annotation),
"bytes": args.annotation.stat().st_size,
"counts": counts,
"dangling_references": dict(dangling),
"valid": not errors,
"errors": errors,
}
args.output.parent.mkdir(parents=True, exist_ok=True)
args.output.write_text(json.dumps(result, indent=2) + "\n", encoding="utf-8")
if errors:
raise SystemExit("; ".join(errors[:5]))
if __name__ == "__main__":
main()
|