Jens/geointel
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Bound evaluator source batches

Browse Source
This commit is contained in:
Jens
2026-07-27 14:52:29 +02:00
parent 31596d98f2
commit ed3f1de201
1 changed files with 22 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files
scripts/evaluate_belgium_building_candidate.py
+22 -15
View File
@@ -180,6 +180,8 @@ def main() -> int:
parser.add_argument("--proposal-crop-scale", type=float, default=1.4) parser.add_argument("--proposal-crop-scale", type=float, default=1.4)
parser.add_argument("--proposal-classifier-batch", type=int, default=64) parser.add_argument("--proposal-classifier-batch", type=int, default=64)
args = parser.parse_args() args = parser.parse_args()
if args.batch < 1:
parser.error("--batch must be positive")
if args.proposal_classifier_batch < 1: if args.proposal_classifier_batch < 1:
parser.error("--proposal-classifier-batch must be positive") parser.error("--proposal-classifier-batch must be positive")
if not 0.0 < args.nms_iou < 1.0: if not 0.0 < args.nms_iou < 1.0:
@@ -211,23 +213,28 @@ def main() -> int:
} }
tiles = [item for item in summary["tiles"] if item.get("kept", True) and item["split"] == args.split] tiles = [item for item in summary["tiles"] if item.get("kept", True) and item["split"] == args.split]
image_paths = [item["image_path"] for item in tiles] image_paths = [item["image_path"] for item in tiles]
results = YOLO(str(args.model)).predict( def predict_bounded(model: YOLO) -> list[Any]:
image_paths, bounded_results: list[Any] = []
conf=min(args.thresholds), for start in range(0, len(image_paths), args.batch):
device=args.device, bounded_results.extend(
augment=args.augment, model.predict(
imgsz=args.imgsz, image_paths[start : start + args.batch],
batch=args.batch, conf=min(args.thresholds),
max_det=args.max_det, device=args.device,
iou=args.nms_iou, augment=args.augment,
verbose=False, imgsz=args.imgsz,
) batch=args.batch,
max_det=args.max_det,
iou=args.nms_iou,
verbose=False,
)
)
return bounded_results
results = predict_bounded(YOLO(str(args.model)))
additional_results = None additional_results = None
if args.additional_model: if args.additional_model:
additional_results = YOLO(str(args.additional_model)).predict( additional_results = predict_bounded(YOLO(str(args.additional_model)))
image_paths, conf=min(args.thresholds), device=args.device, augment=args.augment,
imgsz=args.imgsz, batch=args.batch, max_det=args.max_det, iou=args.nms_iou, verbose=False,
)
observations: list[dict[str, Any]] = [] observations: list[dict[str, Any]] = []
for result_index, (tile, result) in enumerate(zip(tiles, results, strict=True)): for result_index, (tile, result) in enumerate(zip(tiles, results, strict=True)):
height, width = result.orig_shape height, width = result.orig_shape