mxnet.npx.multibox_target¶
-
multibox_target
(anchor=None, label=None, cls_pred=None, overlap_threshold=_Null, ignore_label=_Null, negative_mining_ratio=_Null, negative_mining_thresh=_Null, minimum_negative_samples=_Null, variances=_Null, out=None, name=None, **kwargs)¶ Compute Multibox training targets
- Parameters
anchor (ndarray) – Generated anchor boxes.
label (ndarray) – Object detection labels.
cls_pred (ndarray) – Class predictions.
overlap_threshold (float, optional, default=0.5) – Anchor-GT overlap threshold to be regarded as a positive match.
ignore_label (float, optional, default=-1) – Label for ignored anchors.
negative_mining_ratio (float, optional, default=-1) – Max negative to positive samples ratio, use -1 to disable mining
negative_mining_thresh (float, optional, default=0.5) – Threshold used for negative mining.
minimum_negative_samples (int, optional, default='0') – Minimum number of negative samples.
variances (tuple of <float>, optional, default=[0.1,0.1,0.2,0.2]) – Variances to be encoded in box regression target.
out (ndarray, optional) – The output ndarray to hold the result.
- Returns
out – The output of this function.
- Return type
ndarray or list of ndarrays