Package ai.djl.nn.convolutional

Class Conv1d

java.lang.Object

ai.djl.nn.AbstractBaseBlock

ai.djl.nn.AbstractBlock

ai.djl.nn.convolutional.Convolution

ai.djl.nn.convolutional.Conv1d

All Implemented Interfaces:

Block

public class Conv1d
extends Convolution

A Conv1d layer works similar to Convolution layer with the exception of the number of dimension it operates on being only one, which is LayoutType.WIDTH. The channel of the input data may be more than one, depending on what data is processed. Each filter slides through the data with only one direction of movement along the dimension itself.

Commonly, this kind of convolution layer, as proposed in this paper is used in tasks utilizing serial data, enabling convolutional processing of 1-dimensional data such as time-series data (stock price, weather, ECG) and text/speech data without the need of transforming it to 2-dimensional data to be processed by Conv2d, though this is quite a common technique as well.

The input to a Conv1d is an NDList with a single 3-D NDArray. The layout of the NDArray must be "NCW". The shapes are

• data: (batch_size, channel, width)
• weight: (num_filter, channel, kernel[0])
• bias: (num_filter,)
• out: (batch_size, num_filter, out_width)
out_width = f(width, kernel[0], pad[0], stride[0], dilate[0])
where f(x, k, p, s, d) = floor((x + 2 * p - d * (k - 1) - 1)/s) + 1

Both weight and bias are learn-able parameters.

See Also:

• Convolution

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static final class	Conv1d.Builder	The Builder to construct a Conv1d type of Block.

Nested classes/interfaces inherited from class ai.djl.nn.convolutional.Convolution

Convolution.ConvolutionBuilder<T extends Convolution.ConvolutionBuilder>

Field Summary

Fields inherited from class ai.djl.nn.convolutional.Convolution

bias, dilation, filters, groups, includeBias, kernelShape, padding, stride, weight

Fields inherited from class ai.djl.nn.AbstractBlock

children, parameters

Fields inherited from class ai.djl.nn.AbstractBaseBlock

inputNames, inputShapes, outputDataTypes, version

Method Summary

Modifier and Type	Method	Description
static Conv1d.Builder	builder()	Creates a builder to build a Conv1d.
static NDList	conv1d(NDArray input, NDArray weight)	Applies 1D convolution over an input signal composed of several input planes.
static NDList	conv1d(NDArray input, NDArray weight, NDArray bias)	Applies 1D convolution over an input signal composed of several input planes.
static NDList	conv1d(NDArray input, NDArray weight, NDArray bias, Shape stride)	Applies 1D convolution over an input signal composed of several input planes.
static NDList	conv1d(NDArray input, NDArray weight, NDArray bias, Shape stride, Shape padding)	Applies 1D convolution over an input signal composed of several input planes.
static NDList	conv1d(NDArray input, NDArray weight, NDArray bias, Shape stride, Shape padding, Shape dilation)	Applies 1D convolution over an input signal composed of several input planes.
static NDList	conv1d(NDArray input, NDArray weight, NDArray bias, Shape stride, Shape padding, Shape dilation, int groups)	Applies 1D convolution over an input signal composed of several input planes.
protected LayoutType[]	getExpectedLayout()	Returns the expected layout of the input.
protected String	getStringLayout()	Returns the string representing the layout of the input.
protected int	numDimensions()	Returns the number of dimensions of the input.

Methods inherited from class ai.djl.nn.convolutional.Convolution

beforeInitialize, forwardInternal, getDilation, getFilters, getGroups, getKernelShape, getOutputShapes, getPadding, getStride, isIncludeBias, loadMetadata, prepare

Methods inherited from class ai.djl.nn.AbstractBlock

addChildBlock, addChildBlock, addChildBlockSingleton, addParameter, getChildren, getDirectParameters

Methods inherited from class ai.djl.nn.AbstractBaseBlock

cast, clear, describeInput, forward, forward, forwardInternal, getInputShapes, getOutputDataTypes, getParameters, initialize, initializeChildBlocks, isInitialized, loadParameters, readInputShapes, saveInputShapes, saveMetadata, saveParameters, setInitializer, setInitializer, setInitializer, toString

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Methods inherited from interface ai.djl.nn.Block

forward, freezeParameters, freezeParameters, getOutputShapes

Method Details

getExpectedLayout

protected LayoutType[] getExpectedLayout()

Returns the expected layout of the input.

Specified by:

getExpectedLayout in class Convolution

Returns:

the expected layout of the input

getStringLayout

protected String getStringLayout()

Returns the string representing the layout of the input.

Specified by:

getStringLayout in class Convolution

Returns:

the string representing the layout of the input

numDimensions

protected int numDimensions()

Returns the number of dimensions of the input.

Specified by:

numDimensions in class Convolution

Returns:

the number of dimensions of the input

conv1d

public static NDList conv1d(NDArray input,
 NDArray weight)

Applies 1D convolution over an input signal composed of several input planes.

Parameters:

input - the input NDArray of shape (batchSize, inputChannel, width)

weight - filters NDArray of shape (outChannel, inputChannel/groups, width)

Returns:

the output of the conv1d operation

conv1d

public static NDList conv1d(NDArray input,
 NDArray weight,
 NDArray bias)

Applies 1D convolution over an input signal composed of several input planes.

Parameters:

input - the input NDArray of shape (batchSize, inputChannel, width)

weight - filters NDArray of shape (outChannel, inputChannel/groups, width)

bias - bias NDArray of shape (outChannel)

Returns:

the output of the conv1d operation

conv1d

public static NDList conv1d(NDArray input,
 NDArray weight,
 NDArray bias,
 Shape stride)

Applies 1D convolution over an input signal composed of several input planes.

Parameters:

input - the input NDArray of shape (batchSize, inputChannel, width)

weight - filters NDArray of shape (outChannel, inputChannel/groups, width)

bias - bias NDArray of shape (outChannel)

stride - the stride of the convolving kernel: Shape(width)

Returns:

the output of the conv1d operation

conv1d

public static NDList conv1d(NDArray input,
 NDArray weight,
 NDArray bias,
 Shape stride,
 Shape padding)

Applies 1D convolution over an input signal composed of several input planes.

Parameters:

input - the input NDArray of shape (batchSize, inputChannel, width)

weight - filters NDArray of shape (outChannel, inputChannel/groups, width)

bias - bias NDArray of shape (outChannel)

stride - the stride of the convolving kernel: Shape(width)

padding - implicit paddings on both sides of the input: Shape(width)

Returns:

the output of the conv1d operation

conv1d

public static NDList conv1d(NDArray input,
 NDArray weight,
 NDArray bias,
 Shape stride,
 Shape padding,
 Shape dilation)

Applies 1D convolution over an input signal composed of several input planes.

Parameters:

input - the input NDArray of shape (batchSize, inputChannel, width)

weight - filters NDArray of shape (outChannel, inputChannel/groups, width)

bias - bias NDArray of shape (outChannel)

stride - the stride of the convolving kernel: Shape(width)

padding - implicit paddings on both sides of the input: Shape(width)

dilation - the spacing between kernel elements: Shape(width)

Returns:

the output of the conv1d operation

conv1d

public static NDList conv1d(NDArray input,
 NDArray weight,
 NDArray bias,
 Shape stride,
 Shape padding,
 Shape dilation,
 int groups)

Applies 1D convolution over an input signal composed of several input planes.

Parameters:

input - the input NDArray of shape (batchSize, inputChannel, width)

weight - filters NDArray of shape (outChannel, inputChannel/groups, width)

bias - bias NDArray of shape (outChannel)

stride - the stride of the convolving kernel: Shape(width)

padding - implicit paddings on both sides of the input: Shape(width)

dilation - the spacing between kernel elements: Shape(width)

groups - split input into groups: input channel(input.size(1)) should be divisible by the number of groups

Returns:

the output of the conv1d operation

builder

public static Conv1d.Builder builder()

Creates a builder to build a Conv1d.

Returns:

a new builder