open3d::core::nns::FaissIndex Class Reference

Faiss for nearest neighbor search. More...

#include <FaissIndex.h>

Inheritance diagram for open3d::core::nns::FaissIndex:

Public Member Functions
	FaissIndex ()
	Default Constructor. More...
	FaissIndex (const Tensor &dataset_points)
	Parameterized Constructor. More...
	~FaissIndex ()
	FaissIndex (const FaissIndex &)=delete
FaissIndex &	operator= (const FaissIndex &)=delete
bool	SetTensorData (const Tensor &dataset_points) override
bool	SetTensorData (const Tensor &dataset_points, double radius) override
std::pair< Tensor, Tensor >	SearchKnn (const Tensor &query_points, int knn) const override
std::tuple< Tensor, Tensor, Tensor >	SearchRadius (const Tensor &query_points, const Tensor &radii, bool sort) const override
std::tuple< Tensor, Tensor, Tensor >	SearchRadius (const Tensor &query_points, double radius, bool sort) const override
std::tuple< Tensor, Tensor, Tensor >	SearchHybrid (const Tensor &query_points, double radius, int max_knn) const override
Public Member Functions inherited from open3d::core::nns::NNSIndex
	NNSIndex ()
	Default Constructor. More...
virtual	~NNSIndex ()
	NNSIndex (const NNSIndex &)=delete
NNSIndex &	operator= (const NNSIndex &)=delete
int	GetDimension () const
size_t	GetDatasetSize () const
Dtype	GetDtype () const
Device	GetDevice () const

Protected Attributes
std::unique_ptr< faiss::Index >	index
Protected Attributes inherited from open3d::core::nns::NNSIndex
Tensor	dataset_points_

Detailed Description

Faiss for nearest neighbor search.

Constructor & Destructor Documentation

FaissIndex() [1/3]

open3d::core::nns::FaissIndex::FaissIndex

(

)

Default Constructor.

FaissIndex() [2/3]

open3d::core::nns::FaissIndex::FaissIndex

(

const Tensor &

dataset_points

)

Parameterized Constructor.

Parameters

dataset_points

Provides tensor from which Faiss Index is constructed.

~FaissIndex()

open3d::core::nns::FaissIndex::~FaissIndex

(

)

FaissIndex() [3/3]

open3d::core::nns::FaissIndex::FaissIndex ( const FaissIndex &  )

delete

Member Function Documentation

operator=()

FaissIndex& open3d::core::nns::FaissIndex::operator= ( const FaissIndex &  )

delete

SearchHybrid()

std::tuple< Tensor, Tensor, Tensor > open3d::core::nns::FaissIndex::SearchHybrid ( const Tensor &  query_points, double  radius, int  max_knn  ) const

overridevirtual

Perform hybrid search.

Parameters

query_points	Query points. Must be Float32, 2D, with shape {n, d}.
radius	Radius.
max_knn	Maximum number of neighbor to search per query point.

Returns

Tuple of Tensors, (indices, distances, counts):

• indices: Tensor of shape {n, knn}, with dtype Int32.
• distances: Tensor of shape {n, knn}, with dtype Float32.
• counts: Counts of neighbour for each query points. [Tensor of shape {n}, with dtype Int32].

Implements open3d::core::nns::NNSIndex.

SearchKnn()

std::pair< Tensor, Tensor > open3d::core::nns::FaissIndex::SearchKnn ( const Tensor &  query_points, int  knn  ) const

overridevirtual

Perform K nearest neighbor search.

Parameters

query_points	Query points. Must be Float32, 2D, with shape {n, d}.
knn	Number of nearest neighbor to search.

Returns

Pair of Tensors: (indices, distances):

• indices: Tensor of shape {n, knn}, with dtype Int32.
• distainces: Tensor of shape {n, knn}, same dtype with dataset_points.

Implements open3d::core::nns::NNSIndex.

SearchRadius() [1/2]

std::tuple<Tensor, Tensor, Tensor> open3d::core::nns::FaissIndex::SearchRadius ( const Tensor &  query_points, const Tensor &  radii, bool  sort  ) const

inlineoverridevirtual

Perform radius search with multiple radii.

Parameters

query_points	Query points. Must be Float32, 2D, with shape {n, d}.
radii	list of radius. Must be 1D, with shape {n, }.

Returns

Tuple of Tensors: (indices, distances, counts):

• indicecs: Tensor of shape {total_num_neighbors,}, dtype Int32.
• distances: Tensor of shape {total_num_neighbors,}, same dtype with dataset_points.
• counts: Tensor of shape {n,}, dtype Int32.

Implements open3d::core::nns::NNSIndex.

SearchRadius() [2/2]

std::tuple<Tensor, Tensor, Tensor> open3d::core::nns::FaissIndex::SearchRadius ( const Tensor &  query_points, double  radius, bool  sort  ) const

inlineoverridevirtual

Perform radius search.

Parameters

query_points	Query points. Must be Float32, 2D, with shape {n, d}.
radius	Radius.

Returns

Tuple of Tensors, (indices, distances, counts):

• indicecs: Tensor of shape {total_num_neighbors,}, dtype Int32.
• distances: Tensor of shape {total_num_neighbors,}, same dtype with dataset_points.
• counts: Tensor of shape {n,}, dtype Int32.

Implements open3d::core::nns::NNSIndex.

SetTensorData() [1/2]

bool open3d::core::nns::FaissIndex::SetTensorData ( const Tensor &  dataset_points )

overridevirtual

Set the data for the nearest neighbor search.

Parameters

dataset_points

Dataset points for KDTree construction. Must be 2D, with shape {n, d}.

Returns

Returns true if the construction success, otherwise false.

Implements open3d::core::nns::NNSIndex.

SetTensorData() [2/2]

bool open3d::core::nns::FaissIndex::SetTensorData ( const Tensor &  dataset_points, double  radius  )

inlineoverridevirtual

Set the data for the nearest neighbor search.

Parameters

dataset_points

Dataset points for KDTree construction. Must be 2D, with shape {n, d}.

Returns

Returns true if the construction success, otherwise false.

Implements open3d::core::nns::NNSIndex.

Field Documentation

index

std::unique_ptr<faiss::Index> open3d::core::nns::FaissIndex::index

protected

---

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

• cpp/open3d/core/nns/FaissIndex.h
• cpp/open3d/core/nns/FaissIndex.cpp