#include "NvInfer.h"

Classes
class	nvinfer1::plugin::INvPlugin
	Common interface for the Nvidia created plugins. More...

struct	nvinfer1::plugin::Quadruple
	The Permute plugin layer permutes the input tensor by changing the memory order of the data. Quadruple defines a structure that contains an array of 4 integers. They can represent the permute orders or the strides in each dimension. More...

struct	nvinfer1::plugin::PriorBoxParameters
	The PriorBox plugin layer generates the prior boxes of designated sizes and aspect ratios across all dimensions (H x W). PriorBoxParameters defines a set of parameters for creating the PriorBox plugin layer. It contains: More...

struct	nvinfer1::plugin::GridAnchorParameters
	The Anchor Generator plugin layer generates the prior boxes of designated sizes and aspect ratios across all dimensions (H x W). GridAnchorParameters defines a set of parameters for creating the plugin layer for all feature maps. It contains: More...

struct	nvinfer1::plugin::DetectionOutputParameters
	The DetectionOutput plugin layer generates the detection output based on location and confidence predictions by doing non maximum suppression. This plugin first decodes the bounding boxes based on the anchors generated. It then performs non_max_suppression on the decoded bouding boxes. DetectionOutputParameters defines a set of parameters for creating the DetectionOutput plugin layer. It contains: More...

struct	nvinfer1::plugin::softmaxTree
	The Region plugin layer performs region proposal calculation: generate 5 bounding boxes per cell (for yolo9000, generate 3 bounding boxes per cell). For each box, calculating its probablities of objects detections from 80 pre-defined classifications (yolo9000 has 9416 pre-defined classifications, and these 9416 items are organized as work-tree structure). RegionParameters defines a set of parameters for creating the Region plugin layer. More...

struct	nvinfer1::plugin::RegionParameters

struct	nvinfer1::plugin::NMSParameters
	The NMSParameters are used by the BatchedNMSPlugin for performing the non_max_suppression operation over boxes for object detection networks. More...

Namespaces
	nvinfer1
	The TensorRT API version 1 namespace.

Enumerations
enum	nvinfer1::PluginType : int { nvinfer1::PluginType::kFASTERRCNN = 0, nvinfer1::PluginType::kNORMALIZE = 1, nvinfer1::PluginType::kPERMUTE = 2, nvinfer1::PluginType::kPRIORBOX = 3, nvinfer1::PluginType::kSSDDETECTIONOUTPUT = 4, nvinfer1::PluginType::kCONCAT = 5, nvinfer1::PluginType::kPRELU = 6, nvinfer1::PluginType::kYOLOREORG = 7, nvinfer1::PluginType::kYOLOREGION = 8, nvinfer1::PluginType::kANCHORGENERATOR = 9 }
	The type values for the various plugins. More...

enum	CodeTypeSSD : int { nvinfer1::plugin::CodeTypeSSD::CORNER = 0, nvinfer1::plugin::CodeTypeSSD::CENTER_SIZE = 1, nvinfer1::plugin::CodeTypeSSD::CORNER_SIZE = 2, nvinfer1::plugin::CodeTypeSSD::TF_CENTER = 3 }
	The type of encoding used for decoding the bounding boxes and loc_data.

Functions
template<>
int	nvinfer1::EnumMax< PluginType > ()

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createFasterRCNNPlugin (int featureStride, int preNmsTop, int nmsMaxOut, float iouThreshold, float minBoxSize, float spatialScale, DimsHW pooling, Weights anchorRatios, Weights anchorScales)
	Create a plugin layer that fuses the RPN and ROI pooling using user-defined parameters. More...

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createFasterRCNNPlugin (const void *data, size_t length)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createSSDNormalizePlugin (const Weights *scales, bool acrossSpatial, bool channelShared, float eps)
	The Normalize plugin layer normalizes the input to have L2 norm of 1 with scale learnable. More...

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createSSDNormalizePlugin (const void *data, size_t length)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createSSDPermutePlugin (Quadruple permuteOrder)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createSSDPermutePlugin (const void *data, size_t length)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createSSDPriorBoxPlugin (PriorBoxParameters param)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createSSDPriorBoxPlugin (const void *data, size_t length)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createSSDAnchorGeneratorPlugin (GridAnchorParameters *param, int numLayers)
	The Grid Anchor Generator plugin layer generates the prior boxes of designated sizes and aspect ratios across all dimensions (H x W) for all feature maps. GridAnchorParameters defines a set of parameters for creating the GridAnchorGenerator plugin layer. More...

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createSSDAnchorGeneratorPlugin (const void *data, size_t length)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createSSDDetectionOutputPlugin (DetectionOutputParameters param)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createSSDDetectionOutputPlugin (const void *data, size_t length)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createConcatPlugin (int concatAxis, bool ignoreBatch)
	The Concat plugin layer basically performs the concatention for 4D tensors. Unlike the Concatenation layer in early version of TensorRT, it allows the user to specify the axis along which to concatenate. The axis can be 1 (across channel), 2 (across H), or 3 (across W). More particularly, this Concat plugin layer also implements the "ignoring the batch dimension" switch. If turned on, all the input tensors will be treated as if their batch sizes were 1. More...

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createConcatPlugin (const void *data, size_t length)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createPReLUPlugin (float negSlope)
	The PReLu plugin layer performs leaky ReLU for 4D tensors. Give an input value x, the PReLU layer computes the output as x if x > 0 and negative_slope //! x if x <= 0. More...

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createPReLUPlugin (const void *data, size_t length)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createYOLOReorgPlugin (int stride)
	The Reorg plugin layer maps the 512x26x26 feature map onto a 2048x13x13 feature map, so that it can be concatenated with the feature maps at 13x13 resolution. More...

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createYOLOReorgPlugin (const void *data, size_t length)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createYOLORegionPlugin (RegionParameters params)

TENSORRTAPI INvPlugin *	nvinfer1::plugin::createYOLORegionPlugin (const void *data, size_t length)

TENSORRTAPI nvinfer1::IPluginV2 *	createRPNROIPlugin (int featureStride, int preNmsTop, int nmsMaxOut, float iouThreshold, float minBoxSize, float spatialScale, nvinfer1::DimsHW pooling, nvinfer1::Weights anchorRatios, nvinfer1::Weights anchorScales)
	Create a plugin layer that fuses the RPN and ROI pooling using user-defined parameters. Registered plugin type "RPROI_TRT". Registered plugin version "1". More...

TENSORRTAPI nvinfer1::IPluginV2 *	createNormalizePlugin (const nvinfer1::Weights *scales, bool acrossSpatial, bool channelShared, float eps)
	The Normalize plugin layer normalizes the input to have L2 norm of 1 with scale learnable. Registered plugin type "Normalize_TRT". Registered plugin version "1". More...

TENSORRTAPI nvinfer1::IPluginV2 *	createPriorBoxPlugin (nvinfer1::plugin::PriorBoxParameters param)
	The PriorBox plugin layer generates the prior boxes of designated sizes and aspect ratios across all dimensions (H x W). PriorBoxParameters defines a set of parameters for creating the PriorBox plugin layer. Registered plugin type "PriorBox_TRT". Registered plugin version "1".

TENSORRTAPI nvinfer1::IPluginV2 *	createAnchorGeneratorPlugin (nvinfer1::plugin::GridAnchorParameters *param, int numLayers)
	The Grid Anchor Generator plugin layer generates the prior boxes of designated sizes and aspect ratios across all dimensions (H x W) for all feature maps. GridAnchorParameters defines a set of parameters for creating the GridAnchorGenerator plugin layer. Registered plugin type "GridAnchor_TRT". Registered plugin version "1".

TENSORRTAPI nvinfer1::IPluginV2 *	createNMSPlugin (nvinfer1::plugin::DetectionOutputParameters param)
	The DetectionOutput plugin layer generates the detection output based on location and confidence predictions by doing non maximum suppression. DetectionOutputParameters defines a set of parameters for creating the DetectionOutput plugin layer. Registered plugin type "NMS_TRT". Registered plugin version "1".

TENSORRTAPI nvinfer1::IPluginV2 *	createLReLUPlugin (float negSlope)
	The LReLu plugin layer performs leaky ReLU for 4D tensors. Give an input value x, the PReLU layer computes the output as x if x > 0 and negative_slope //! x if x <= 0. Registered plugin type "LReLU_TRT". Registered plugin version "1". More...

TENSORRTAPI nvinfer1::IPluginV2 *	createReorgPlugin (int stride)
	The Reorg plugin reshapes input of shape CxHxW into a (Cstridestride)x(H/stride)x(W/stride) shape, used in YOLOv2. It does that by taking 1 x stride x stride slices from tensor and flattening them into (stridexstride) x 1 x 1 shape. Registered plugin type "Reorg_TRT". Registered plugin version "1". More...

TENSORRTAPI nvinfer1::IPluginV2 *	createRegionPlugin (nvinfer1::plugin::RegionParameters params)
	The Region plugin layer performs region proposal calculation: generate 5 bounding boxes per cell (for yolo9000, generate 3 bounding boxes per cell). For each box, calculating its probablities of objects detections from 80 pre-defined classifications (yolo9000 has 9416 pre-defined classifications, and these 9416 items are organized as work-tree structure). RegionParameters defines a set of parameters for creating the Region plugin layer. Registered plugin type "Region_TRT". Registered plugin version "1".

TENSORRTAPI nvinfer1::IPluginV2 *	createClipPlugin (const char *layerName, float clipMin, float clipMax)
	The Clip Plugin performs a clip operation on the input tensor. It clips the tensor values to a specified min and max. Any value less than clipMin are set to clipMin. Any values greater than clipMax are set to clipMax. For example, this plugin can be used to perform a Relu6 operation by specifying clipMin=0.0 and clipMax=6.0 Registered plugin type "Clip_TRT". Registered plugin version "1". More...

TENSORRTAPI nvinfer1::IPluginV2 *	createBatchedNMSPlugin (nvinfer1::plugin::NMSParameters param)
	The BatchedNMS Plugin performs non_max_suppression on the input boxes, per batch, across all classes. It greedily selects a subset of bounding boxes in descending order of score. Prunes away boxes that have a high intersection-over-union (IOU) overlap with previously selected boxes. Bounding boxes are supplied as [y1, x1, y2, x2], where (y1, x1) and (y2, x2) are the coordinates of any diagonal pair of box corners and the coordinates can be provided as normalized (i.e., lying in the interval [0, 1]) or absolute. The plugin expects two inputs. Input0 is expected to be 4-D float boxes tensor of shape [batch_size, num_boxes, q, 4], where q can be either 1 (if shareLocation is true) or num_classes. Input1 is expected to be a 3-D float scores tensor of shape [batch_size, num_boxes, num_classes] representing a single score corresponding to each box. The plugin returns four outputs. num_detections : A [batch_size] int32 tensor indicating the number of valid detections per batch item. Can be less than keepTopK. Only the top num_detections[i] entries in nmsed_boxes[i], nmsed_scores[i] and nmsed_classes[i] are valid. nmsed_boxes : A [batch_size, max_detections, 4] float32 tensor containing the co-ordinates of non-max suppressed boxes. nmsed_scores : A [batch_size, max_detections] float32 tensor containing the scores for the boxes. nmsed_classes : A [batch_size, max_detections] float32 tensor containing the classes for the boxes. More...

TENSORRTAPI bool	initLibNvInferPlugins (void logger, const char libNamespace)
	Initialize and register all the existing TensorRT plugins to the Plugin Registry with an optional namespace. The plugin library author should ensure that this function name is unique to the library. This function should be called once before accessing the Plugin Registry. More...

Detailed Description

This is the API for the Nvidia provided TensorRT plugins.

Function Documentation

TENSORRTAPI nvinfer1::IPluginV2* createBatchedNMSPlugin ( nvinfer1::plugin::NMSParameters param )

The BatchedNMS Plugin performs non_max_suppression on the input boxes, per batch, across all classes. It greedily selects a subset of bounding boxes in descending order of score. Prunes away boxes that have a high intersection-over-union (IOU) overlap with previously selected boxes. Bounding boxes are supplied as [y1, x1, y2, x2], where (y1, x1) and (y2, x2) are the coordinates of any diagonal pair of box corners and the coordinates can be provided as normalized (i.e., lying in the interval [0, 1]) or absolute. The plugin expects two inputs. Input0 is expected to be 4-D float boxes tensor of shape [batch_size, num_boxes, q, 4], where q can be either 1 (if shareLocation is true) or num_classes. Input1 is expected to be a 3-D float scores tensor of shape [batch_size, num_boxes, num_classes] representing a single score corresponding to each box. The plugin returns four outputs. num_detections : A [batch_size] int32 tensor indicating the number of valid detections per batch item. Can be less than keepTopK. Only the top num_detections[i] entries in nmsed_boxes[i], nmsed_scores[i] and nmsed_classes[i] are valid. nmsed_boxes : A [batch_size, max_detections, 4] float32 tensor containing the co-ordinates of non-max suppressed boxes. nmsed_scores : A [batch_size, max_detections] float32 tensor containing the scores for the boxes. nmsed_classes : A [batch_size, max_detections] float32 tensor containing the classes for the boxes.

Registered plugin type "BatchedNMS_TRT". Registered plugin version "1".

TENSORRTAPI nvinfer1::IPluginV2* createClipPlugin	(	const char *	layerName,
		float	clipMin,
		float	clipMax
	)

The Clip Plugin performs a clip operation on the input tensor. It clips the tensor values to a specified min and max. Any value less than clipMin are set to clipMin. Any values greater than clipMax are set to clipMax. For example, this plugin can be used to perform a Relu6 operation by specifying clipMin=0.0 and clipMax=6.0 Registered plugin type "Clip_TRT". Registered plugin version "1".

Parameters

layerName	The name of the TensorRT layer.
clipMin	The minimum value to clip to.
clipMax	The maximum value to clip to.

TENSORRTAPI nvinfer1::IPluginV2* createLReLUPlugin ( float negSlope )

The LReLu plugin layer performs leaky ReLU for 4D tensors. Give an input value x, the PReLU layer computes the output as x if x > 0 and negative_slope //! x if x <= 0. Registered plugin type "LReLU_TRT". Registered plugin version "1".

Parameters

negSlope Negative_slope value.

TENSORRTAPI nvinfer1::IPluginV2* createNormalizePlugin	(	const nvinfer1::Weights *	scales,
		bool	acrossSpatial,
		bool	channelShared,
		float	eps
	)

The Normalize plugin layer normalizes the input to have L2 norm of 1 with scale learnable. Registered plugin type "Normalize_TRT". Registered plugin version "1".

Parameters

scales	Scale weights that are applied to the output tensor.
acrossSpatial	Whether to compute the norm over adjacent channels (acrossSpatial is true) or nearby spatial locations (within channel in which case acrossSpatial is false).
channelShared	Whether the scale weight(s) is shared across channels.
eps	Epsilon for not diviiding by zero.

TENSORRTAPI nvinfer1::IPluginV2* createReorgPlugin ( int stride )

The Reorg plugin reshapes input of shape CxHxW into a (C*stride*stride)x(H/stride)x(W/stride) shape, used in YOLOv2. It does that by taking 1 x stride x stride slices from tensor and flattening them into (stridexstride) x 1 x 1 shape. Registered plugin type "Reorg_TRT". Registered plugin version "1".

Parameters

stride Strides in H and W, it should divide both H and W. Also stride * stride should be less than or equal to C.

TENSORRTAPI nvinfer1::IPluginV2* createRPNROIPlugin	(	int	featureStride,
		int	preNmsTop,
		int	nmsMaxOut,
		float	iouThreshold,
		float	minBoxSize,
		float	spatialScale,
		nvinfer1::DimsHW	pooling,
		nvinfer1::Weights	anchorRatios,
		nvinfer1::Weights	anchorScales
	)

Create a plugin layer that fuses the RPN and ROI pooling using user-defined parameters. Registered plugin type "RPROI_TRT". Registered plugin version "1".

Parameters

featureStride	Feature stride.
preNmsTop	Number of proposals to keep before applying NMS.
nmsMaxOut	Number of remaining proposals after applying NMS.
iouThreshold	IoU threshold.
minBoxSize	Minimum allowed bounding box size before scaling.
spatialScale	Spatial scale between the input image and the last feature map.
pooling	Spatial dimensions of pooled ROIs.
anchorRatios	Aspect ratios for generating anchor windows.
anchorScales	Scales for generating anchor windows.

Returns: Returns a FasterRCNN fused RPN+ROI pooling plugin. Returns nullptr on invalid inputs.

TENSORRTAPI bool initLibNvInferPlugins	(	void *	logger,
		const char *	libNamespace
	)

Initialize and register all the existing TensorRT plugins to the Plugin Registry with an optional namespace. The plugin library author should ensure that this function name is unique to the library. This function should be called once before accessing the Plugin Registry.

Parameters

logger	Logger object to print plugin registration information
libNamespace	Namespace used to register all the plugins in this library

Classes

Namespaces

Enumerations

Functions

Detailed Description

Function Documentation