- nvidia.dali.fn.box_encoder(*inputs, **kwargs)¶
Encodes the input bounding boxes and labels using a set of default boxes (anchors) passed as an argument.
This operator follows the algorithm described in “SSD: Single Shot MultiBox Detector” and implemented in https://github.com/mlperf/training/tree/master/single_stage_detector/ssd. Inputs must be supplied as the following Tensors:
BBoxesthat contain bounding boxes that are represented as
Labelsthat contain the corresponding label for each bounding box.
The results are two tensors:
EncodedBBoxesthat contain M-encoded bounding boxes as
[l,t,r,b], where M is number of anchors.
EncodedLabelsthat contain the corresponding label for each encoded box.
- Supported backends
input0 (TensorList) – Input to the operator.
input1 (TensorList) – Input to the operator.
- Keyword Arguments:
anchors (float or list of float) – Anchors to be used for encoding, as the list of floats is in the
bytes_per_sample_hint (int or list of int, optional, default = ) –
Output size hint, in bytes per sample.
If specified, the operator’s outputs residing in GPU or page-locked host memory will be preallocated to accommodate a batch of samples of this size.
criteria (float, optional, default = 0.5) –
Threshold IoU for matching bounding boxes with anchors.
The value needs to be between 0 and 1.
means (float or list of float, optional, default = [0.0, 0.0, 0.0, 0.0]) – [x y w h] mean values for normalization.
offset (bool, optional, default = False) – Returns normalized offsets
((encoded_bboxes*scale - anchors*scale) - mean) / stdsin EncodedBBoxes that use
preserve (bool, optional, default = False) – Prevents the operator from being removed from the graph even if its outputs are not used.
scale (float, optional, default = 1.0) – Rescales the box and anchor values before the offset is calculated (for example, to return to the absolute values).
seed (int, optional, default = -1) –
If not provided, it will be populated based on the global seed of the pipeline.
stds (float or list of float, optional, default = [1.0, 1.0, 1.0, 1.0]) – [x y w h] standard deviations for offset normalization.