Definition of Non Maximal Supression. More...

Public Member Functions
	NMS ()
	Default constructor required to satisfy the Metric policy. More...

template < typename Archive >
void	serialize (Archive &ar, const uint32_t)
	Serialize the metric. More...

Static Public Member Functions
template < typename BoundingBoxesType , typename ConfidenceScoreType , typename OutputType >
static void	Evaluate (const BoundingBoxesType &boundingBoxes, const ConfidenceScoreType &confidenceScores, OutputType &selectedIndices, const double threshold=0.5)
	Performs non-maximal suppression. More...

Static Public Attributes
static const bool	useCoordinates = UseCoordinates

Detailed Description

template<bool UseCoordinates = false>
class mlpack::metric::NMS< UseCoordinates >

Definition of Non Maximal Supression.

Performs non-maximal suppression (NMS) on the boxes according to their Intersection-over-Union (IoU). NMS iteratively removes lower scoring boxes which have an IoU greater than threshold with another high scoring box.

For bounding box representation there are two common representation either as coordinates i.e. each value in vector represents a coordinate in the format x0, y0, x1, y1 where x0, y0 represent the lower left coordinate and x1, y1, represent upper right coordinate.

Second representation follows the following representation : x0, y0, h, w. Where x0 and y0 are bottom left bounding box coordinates and h, w are height and width of the bounding box.

Template Parameters

UseCoordinates Toggles between the two representation of bounding box. If true, each value in vector represents a coordinate in the formate x0, y0, x1, y1. Else the bounding box is represented as x0, y0, h, w.

Definition at line 40 of file non_maximal_supression.hpp.

Constructor & Destructor Documentation

◆ NMS()

NMS ( )

inline

Default constructor required to satisfy the Metric policy.

Definition at line 44 of file non_maximal_supression.hpp.

References NMS< UseCoordinates >::Evaluate().

Member Function Documentation

◆ Evaluate()

static void Evaluate	(	const BoundingBoxesType &	boundingBoxes,
		const ConfidenceScoreType &	confidenceScores,
		OutputType &	selectedIndices,
		const double	threshold = `0.5`
	)

static

Performs non-maximal suppression.

Parameters

boundingBoxes	Column major representation of bounding boxes i.e. Each column corresponds to a different bounding box. Each bounding box should contain 4 points only either {x1, y1, x2, y2} or {x1, y1, h, w} depending on UseCoordinates parameter.
confidenceScores	Vector containing confidence score corresponding to each bounding box.
selectedIndices	Output of Non Maximal Suppression (NMS) is stored here. It contains a list of indices corresponding to bounding boxes in input parameter, sorted in descending order of the confidence scores.
threshold	Threshold used to discard all overlapping bounding boxes that have IoU greater than the threshold.

Referenced by NMS< UseCoordinates >::NMS().

◆ serialize()

void serialize	(	Archive &	ar,
		const uint32_t
	)

Serialize the metric.

Member Data Documentation

◆ useCoordinates

const bool useCoordinates = UseCoordinates

static

Definition at line 73 of file non_maximal_supression.hpp.

The documentation for this class was generated from the following file:

/home/ryan/src/mlpack.org/_src/mlpack-git/src/mlpack/core/metrics/non_maximal_supression.hpp

Public Member Functions

Static Public Member Functions

Static Public Attributes