NEConv3D Class Reference

Basic function to simulate a 3d convolution. More...

#include <NEConv3D.h>

Collaboration diagram for NEConv3D:

Public Member Functions
	NEConv3D ()
	Constructor. More...
	NEConv3D (const NEConv3D &)=delete
	Prevent instances of this class from being copied (As this class contains pointers) More...
NEConv3D &	operator= (const NEConv3D &)=delete
	Prevent instances of this class from being copied (As this class contains pointers) More...
	NEConv3D (NEConv3D &&)=default
	Default move constructor. More...
NEConv3D &	operator= (NEConv3D &&)=default
	Prevent instances of this class from being moved (As this class contains non movable objects) More...
	~NEConv3D ()
	Default destructor. More...
void	configure (ITensor *input, const ITensor *weights, const ITensor *biases, ITensor *output, const Conv3dInfo &conv_info)
	Set the input and output tensors. More...
void	run () override
	Run the kernels contained in the function. More...
Public Member Functions inherited from IFunction
virtual	~IFunction ()=default
	Destructor. More...
virtual void	prepare ()
	Prepare the function for executing. More...

Static Public Member Functions
static Status	validate (const ITensorInfo *input, const ITensorInfo *weights, const ITensorInfo *biases, const ITensorInfo *output, const Conv3dInfo &conv_info)
	Static function to check if given info will lead to a valid configuration. More...

Detailed Description

Basic function to simulate a 3d convolution.

This function calls one of the following functions:

1. cpu::CpuDirectConv3d

Definition at line 43 of file NEConv3D.h.

Constructor & Destructor Documentation

NEConv3D() [1/3]

NEConv3D

(

)

Constructor.

Definition at line 43 of file NEConv3D.cpp.

                   : _impl(std::make_unique<Impl>())
{
}

NEConv3D() [2/3]

NEConv3D ( const NEConv3D &  )

delete

Prevent instances of this class from being copied (As this class contains pointers)

NEConv3D() [3/3]

NEConv3D ( NEConv3D &&  )

default

Default move constructor.

~NEConv3D()

~NEConv3D ( )

default

Default destructor.

Member Function Documentation

configure()

void configure	(	ITensor *	input,
		const ITensor *	weights,
		const ITensor *	biases,
		ITensor *	output,
		const Conv3dInfo &	conv_info
	)

Set the input and output tensors.

Valid data layouts:

• NDHWC

Valid data type configurations:

src0	src1	src2	dst
F16	F16	F16	F16
F32	F32	F32	F32
QASYMM8	QASYMM8	S32	QASYMM8
QASYMM8_SIGNED	QASYMM8_SIGNED	S32	QASYMM8_SIGNED

Parameters

[in]	input	Source tensor. 4 lower dimensions represent a single input [IFM, width, height, depth], while every optional dimension from 5 and above represent a batch of inputs.
[in]	weights	Weights tensor. Weights are 5D tensor with dimensions [OFM, IFM, kernel_x, kernel_y, kernel_z].
[in]	biases	Biases tensor. Shared biases supported. Biases are 1D tensor with dimensions [OFM].
[out]	output	Destination tensor. 4 lower dimensions represent a single output [OFM, width, height, depth], while the rest represent batch of outputs.
[in]	conv_info	Contains padding, stride, acitvation information described in Conv3dInfo.

Definition at line 49 of file NEConv3D.cpp.

{
    // Perform validate step
    ARM_COMPUTE_ERROR_ON_NULLPTR(input, weights, output);
    ARM_COMPUTE_ERROR_THROW_ON(cpu::CpuDirectConv3d::validate(
        input->info(), weights->info(), ((biases != nullptr) ? biases->info() : nullptr), output->info(), conv_info));
    ARM_COMPUTE_LOG_PARAMS(input, weights, biases, output, conv_info);
 
    auto f = std::make_unique<cpu::CpuDirectConv3d>();
    f->configure(input->info(), weights->info(), ((biases != nullptr) ? biases->info() : nullptr), output->info(),
                 conv_info);
    _impl->op = std::move(f);
 
    if (_impl->op != nullptr)
    {
        _impl->run_pack = {{ACL_SRC_0, input}, {ACL_SRC_1, weights}, {ACL_SRC_2, biases}, {ACL_DST, output}};
    }
}

References arm_compute::ACL_DST, arm_compute::ACL_SRC_0, arm_compute::ACL_SRC_1, arm_compute::ACL_SRC_2, ARM_COMPUTE_ERROR_ON_NULLPTR, ARM_COMPUTE_ERROR_THROW_ON, ARM_COMPUTE_LOG_PARAMS, arm_compute::test::validation::conv_info, ITensor::info(), arm_compute::test::validation::input, and CpuDirectConv3d::validate().

operator=() [1/2]

NEConv3D& operator= ( const NEConv3D &  )

delete

Prevent instances of this class from being copied (As this class contains pointers)

operator=() [2/2]

NEConv3D& operator= ( NEConv3D &&  )

default

Prevent instances of this class from being moved (As this class contains non movable objects)

run()

void run ( )

overridevirtual

Run the kernels contained in the function.

For CPU kernels:

• Multi-threading is used for the kernels which are parallelisable.
• By default std::thread::hardware_concurrency() threads are used.

Note

CPPScheduler::set_num_threads() can be used to manually set the number of threads

For OpenCL kernels:

• All the kernels are enqueued on the queue associated with CLScheduler.
• The queue is then flushed.

Note

The function will not block until the kernels are executed. It is the user's responsibility to wait.

Will call prepare() on first run if hasn't been done

Implements IFunction.

Definition at line 80 of file NEConv3D.cpp.

{
    if (_impl->op != nullptr)
    {
        _impl->op->run(_impl->run_pack);
    }
}

validate()

Status validate ( const ITensorInfo *  input, const ITensorInfo *  weights, const ITensorInfo *  biases, const ITensorInfo *  output, const Conv3dInfo &  conv_info  )

static

Static function to check if given info will lead to a valid configuration.

Similar to NEConv3D::configure()

Returns

a status

Definition at line 69 of file NEConv3D.cpp.

{
    ARM_COMPUTE_RETURN_ON_ERROR(cpu::CpuDirectConv3d::validate(input, weights, biases, output, conv_info));
 
    return Status{};
}

References ARM_COMPUTE_RETURN_ON_ERROR, arm_compute::test::validation::conv_info, arm_compute::test::validation::input, and CpuDirectConv3d::validate().

---

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

• arm_compute/runtime/NEON/functions/NEConv3D.h
• src/runtime/NEON/functions/NEConv3D.cpp