swirlspy.qpf.dl.operators package¶

Submodules¶

swirlspy.qpf.dl.operators.base_rnn module¶

class swirlspy.qpf.dl.operators.base_rnn.BaseStackRNN(base_rnn_class, stack_num=1, name='BaseStackRNN', residual_connection=True, **kwargs)¶

Bases: object

check_concat(states)¶

concat_to_split(concat_states)¶

flatten_add_layout(states, blocked=False)¶

Parameters: states (list of list or list) –
Returns: ret
Return type: list

init_state_vars()¶

Initial state variable for this cell.

Returns: state_vars – starting states for first RNN step
Return type: nested list of Symbol

reset()¶

split_to_concat(split_states)¶

property state_info¶

property state_postfix¶

to_concat(states)¶

to_split(states)¶

unroll(length, inputs=None, begin_states=None, ret_mid=False)¶

class swirlspy.qpf.dl.operators.base_rnn.MyBaseRNNCell(prefix='MyBaseRNNCell', params=None)¶

Bases: mxnet.rnn.rnn_cell.BaseRNNCell

get_current_states()¶

reset()¶: Reset before re-using the cell for another graph.

unroll(length, inputs=None, begin_state=None, ret_mid=False, input_prefix='', layout='TC', merge_outputs=False)¶

Unroll an RNN cell across time steps.

Parameters

length (int) – number of steps to unroll
inputs (Symbol, list of Symbol, or None) –
if inputs is a single Symbol (usually the output of Embedding symbol), it should have shape (batch_size, length, …) if layout == ‘NTC’, or (length, batch_size, …) if layout == ‘TNC’.

If inputs is a list of symbols (usually output of previous unroll), they should all have shape (batch_size, …).

If inputs is None, Placeholder variables are automatically created.
begin_state (nested list of Symbol) – input states. Created by begin_state() or output state of another cell. Created from begin_state() if None.
input_prefix (str) – prefix for automatically created input placehodlers.
layout (str) – layout of input symbol. Only used if inputs is a single Symbol.
merge_outputs (bool) – if False, return outputs as a list of Symbols. If True, concatenate output across time steps and return a single symbol with shape (batch_size, length, …) if layout == ‘NTC’, or (length, batch_size, …) if layout == ‘TNC’.

Returns

outputs (list of Symbol) – output symbols.
states (Symbol or nested list of Symbol) – has the same structure as begin_state()
mid_info (list of Symbol)

class swirlspy.qpf.dl.operators.base_rnn.MyGRU(num_hidden, zoneout=0.0, act_type='tanh', prefix='gru_', params=None)¶

Bases: swirlspy.qpf.dl.operators.base_rnn.MyBaseRNNCell

GRU cell.

Parameters

num_hidden (int) – number of units in output symbol
prefix (str, default ‘rnn_’) – prefix for name of layers (and name of weight if params is None)
params (RNNParams or None) – container for weight sharing between cells. created if None.

property state_info¶: shape(s) of states

swirlspy.qpf.dl.operators.common module¶

class swirlspy.qpf.dl.operators.common.ConstantOp(data)¶

Bases: mxnet.operator.CustomOp

Implementation of mask on minibatch layer.

backward(req, out_grad, in_data, out_data, in_grad, aux)¶

Backward interface. Can override when creating new operators.

Parameters

req (list of str) – how to assign to in_grad. can be ‘null’, ‘write’, or ‘add’. You can optionally use self.assign(dst, req, src) to handle this.
out_grad (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
in_data (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
out_data (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
in_grad (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
aux (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward

forward(is_train, req, in_data, out_data, aux)¶

Forward interface. Can override when creating new operators.

Parameters

is_train (bool) – whether this is for training
req (list of str) – how to assign to out_data. can be ‘null’, ‘write’, or ‘add’. You can optionally use self.assign(dst, req, src) to handle this.
in_data (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward
out_data (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward
aux (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward

class swirlspy.qpf.dl.operators.common.ConstantOpProp(pkl_data)¶

Bases: mxnet.operator.CustomOpProp

create_operator(ctx, shapes, dtypes)¶: Create an operator that carries out the real computation given the context, input shapes, and input data types.

infer_shape(in_shape)¶

infer_shape interface. Can override when creating new operators.

Parameters

in_shape (list) – List of argument shapes in the same order as declared in list_arguments.

Returns

in_shape (list) – List of argument shapes. Can be modified from in_shape.
out_shape (list) – List of output shapes calculated from in_shape, in the same order as declared in list_outputs.
aux_shape (Optional, list) – List of aux shapes calculated from in_shape, in the same order as declared in list_auxiliary_states.

list_arguments()¶

list_arguments interface. Can override when creating new operators.

Returns: arguments – List of argument blob names.
Return type: list

list_outputs()¶

list_outputs interface. Can override when creating new operators.

Returns: outputs – List of output blob names.
Return type: list

class swirlspy.qpf.dl.operators.common.EntropyMultinomialDist¶

Bases: mxnet.operator.CustomOp

backward(req, out_grad, in_data, out_data, in_grad, aux)¶

Backward interface. Can override when creating new operators.

Parameters

req (list of str) – how to assign to in_grad. can be ‘null’, ‘write’, or ‘add’. You can optionally use self.assign(dst, req, src) to handle this.
out_grad (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
in_data (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
out_data (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
in_grad (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
aux (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward

forward(is_train, req, in_data, out_data, aux)¶

Forward interface. Can override when creating new operators.

Parameters

is_train (bool) – whether this is for training
req (list of str) – how to assign to out_data. can be ‘null’, ‘write’, or ‘add’. You can optionally use self.assign(dst, req, src) to handle this.
in_data (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward
out_data (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward
aux (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward

class swirlspy.qpf.dl.operators.common.EntropyMultinomialDistProp¶

Bases: mxnet.operator.CustomOpProp

create_operator(ctx, shapes, dtypes)¶: Create an operator that carries out the real computation given the context, input shapes, and input data types.

infer_shape(in_shape)¶

infer_shape interface. Can override when creating new operators.

Parameters

in_shape (list) – List of argument shapes in the same order as declared in list_arguments.

Returns

in_shape (list) – List of argument shapes. Can be modified from in_shape.
out_shape (list) – List of output shapes calculated from in_shape, in the same order as declared in list_outputs.
aux_shape (Optional, list) – List of aux shapes calculated from in_shape, in the same order as declared in list_auxiliary_states.

list_arguments()¶

list_arguments interface. Can override when creating new operators.

Returns: arguments – List of argument blob names.
Return type: list

list_outputs()¶

list_outputs interface. Can override when creating new operators.

Returns: outputs – List of output blob names.
Return type: list

class swirlspy.qpf.dl.operators.common.IdentityOp(logging_prefix='identity', input_debug=False, grad_debug=False)¶

Bases: mxnet.operator.CustomOp

backward(req, out_grad, in_data, out_data, in_grad, aux)¶

Backward interface. Can override when creating new operators.

Parameters

req (list of str) – how to assign to in_grad. can be ‘null’, ‘write’, or ‘add’. You can optionally use self.assign(dst, req, src) to handle this.
out_grad (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
in_data (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
out_data (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
in_grad (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
aux (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward

forward(is_train, req, in_data, out_data, aux)¶

Forward interface. Can override when creating new operators.

Parameters

is_train (bool) – whether this is for training
req (list of str) – how to assign to out_data. can be ‘null’, ‘write’, or ‘add’. You can optionally use self.assign(dst, req, src) to handle this.
in_data (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward
out_data (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward
aux (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward

class swirlspy.qpf.dl.operators.common.IdentityOpProp(logging_prefix='identity', input_debug=False, grad_debug=False)¶

Bases: mxnet.operator.CustomOpProp

create_operator(ctx, shapes, dtypes)¶: Create an operator that carries out the real computation given the context, input shapes, and input data types.

infer_shape(in_shape)¶

infer_shape interface. Can override when creating new operators.

Parameters

in_shape (list) – List of argument shapes in the same order as declared in list_arguments.

Returns

in_shape (list) – List of argument shapes. Can be modified from in_shape.
out_shape (list) – List of output shapes calculated from in_shape, in the same order as declared in list_outputs.
aux_shape (Optional, list) – List of aux shapes calculated from in_shape, in the same order as declared in list_auxiliary_states.

list_arguments()¶

list_arguments interface. Can override when creating new operators.

Returns: arguments – List of argument blob names.
Return type: list

list_outputs()¶

list_outputs interface. Can override when creating new operators.

Returns: outputs – List of output blob names.
Return type: list

class swirlspy.qpf.dl.operators.common.LogisticRegressionMaskOutput(ignore_label)¶

Bases: mxnet.operator.CustomOp

backward(req, out_grad, in_data, out_data, in_grad, aux)¶

Backward interface. Can override when creating new operators.

Parameters

req (list of str) – how to assign to in_grad. can be ‘null’, ‘write’, or ‘add’. You can optionally use self.assign(dst, req, src) to handle this.
out_grad (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
in_data (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
out_data (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
in_grad (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
aux (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward

forward(is_train, req, in_data, out_data, aux)¶

Forward interface. Can override when creating new operators.

Parameters

is_train (bool) – whether this is for training
req (list of str) – how to assign to out_data. can be ‘null’, ‘write’, or ‘add’. You can optionally use self.assign(dst, req, src) to handle this.
in_data (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward
out_data (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward
aux (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward

class swirlspy.qpf.dl.operators.common.LogisticRegressionMaskOutputProp(ignore_label)¶

Bases: mxnet.operator.CustomOpProp

create_operator(ctx, shapes, dtypes)¶: Create an operator that carries out the real computation given the context, input shapes, and input data types.

infer_shape(in_shape)¶

infer_shape interface. Can override when creating new operators.

Parameters

in_shape (list) – List of argument shapes in the same order as declared in list_arguments.

Returns

in_shape (list) – List of argument shapes. Can be modified from in_shape.
out_shape (list) – List of output shapes calculated from in_shape, in the same order as declared in list_outputs.
aux_shape (Optional, list) – List of aux shapes calculated from in_shape, in the same order as declared in list_auxiliary_states.

list_arguments()¶

list_arguments interface. Can override when creating new operators.

Returns: arguments – List of argument blob names.
Return type: list

list_outputs()¶

list_outputs interface. Can override when creating new operators.

Returns: outputs – List of output blob names.
Return type: list

class swirlspy.qpf.dl.operators.common.MyConstant(value)¶: Bases: mxnet.initializer.Initializer

class swirlspy.qpf.dl.operators.common.SaveNpyOp(save_name='op', save_dir=None)¶

Bases: mxnet.operator.CustomOp

backward(req, out_grad, in_data, out_data, in_grad, aux)¶

Backward interface. Can override when creating new operators.

Parameters

req (list of str) – how to assign to in_grad. can be ‘null’, ‘write’, or ‘add’. You can optionally use self.assign(dst, req, src) to handle this.
out_grad (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
in_data (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
out_data (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
in_grad (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward
aux (list of NDArrays) – input and output for backward. See document for corresponding arguments of Operator::Backward

forward(is_train, req, in_data, out_data, aux)¶

Forward interface. Can override when creating new operators.

Parameters

is_train (bool) – whether this is for training
req (list of str) – how to assign to out_data. can be ‘null’, ‘write’, or ‘add’. You can optionally use self.assign(dst, req, src) to handle this.
in_data (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward
out_data (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward
aux (list of NDArrays) – input, output, and auxiliary states for forward. See document for corresponding arguments of Operator::Forward

class swirlspy.qpf.dl.operators.common.SaveNpyOpProp(save_name='op', save_dir='.')¶

Bases: mxnet.operator.CustomOpProp

create_operator(ctx, shapes, dtypes)¶: Create an operator that carries out the real computation given the context, input shapes, and input data types.

infer_shape(in_shape)¶

infer_shape interface. Can override when creating new operators.

Parameters

in_shape (list) – List of argument shapes in the same order as declared in list_arguments.

Returns

in_shape (list) – List of argument shapes. Can be modified from in_shape.
out_shape (list) – List of output shapes calculated from in_shape, in the same order as declared in list_outputs.
aux_shape (Optional, list) – List of aux shapes calculated from in_shape, in the same order as declared in list_auxiliary_states.

list_arguments()¶

list_arguments interface. Can override when creating new operators.

Returns: arguments – List of argument blob names.
Return type: list

list_outputs()¶

list_outputs interface. Can override when creating new operators.

Returns: outputs – List of output blob names.
Return type: list

swirlspy.qpf.dl.operators.common.calculate_derivative(data, deriv_variable, step_size, batch_size)¶

Calculate derivatives of input data tensor.

Parameters

data (mx.sym.Symbol : data to take derivative over) – Shape : (seq_len, batch_size, 1, H, W)
deriv_variable (str : 'x' or 'y' or 't') –
step_size (float : step size, e.g. dx, dy or dt (e.g. dt=6min)) –
batch_size (int) –
Returns –
-------- –
deriv (mx.sym.Symbol : tensor of (approximate) derivatives) – Shape : (seq_len, batch_size, 1, H, W)

swirlspy.qpf.dl.operators.common.calculate_lagrangian_derivative(data, motion_vector_field, step_sizes, batch_size)¶

Calculate lagrangian derivative (in 2 dimensions) of input data tensor.

Parameters

data (mx.sym.Symbol : data to take derivative over) – Shape : (seq_len, batch_size, 1, H, W)
motion_vector_field (list of mx.sym.Symbol : velocity fields [u, v]) –
step_sizes (list of float : step sizes [dx, dy, dt]) –
batch_size (int) –
Returns –
-------- –
lagragian_deriv (mx.sym.Symbol :) – Shape : (seq_len, batch_size, 1, H, W)

swirlspy.qpf.dl.operators.common.constant(data, name='constant')¶

swirlspy.qpf.dl.operators.common.entropy_multinomial(data, name='entropy')¶

swirlspy.qpf.dl.operators.common.grid_generator(batch_size, height, width, normalize=True)¶

Generate the grid based on width and height

Parameters

batch_size (int) –
width (int) –
height (int) –
normalize (bool) – Whether to normalize the grid elements into [-1, 1]

Returns

ret – Shape : (batch_size, 2, height, width), the channel contains (x, y)

Return type

mx.sym.Symbol

swirlspy.qpf.dl.operators.common.group_add(lhs, rhs)¶

Parameters

lhs (list of mx.sym.Symbol) –
rhs (list of mx.sym.Symbol) –

Returns

ret

Return type

list of mx.sym.Symbol

swirlspy.qpf.dl.operators.common.identity(data, name='identity', logging_prefix=None, input_debug=False, grad_debug=False)¶

swirlspy.qpf.dl.operators.common.logistic_regression_mask_output(data, label, ignore_label, name=None)¶

swirlspy.qpf.dl.operators.common.masked_gdl_loss(pred, gt, mask)¶

Parameters

pred (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)
gt (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)
mask (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)

Returns

gdl – Shape: (seq_len, batch_size)

Return type

mx.sym.Symbol

swirlspy.qpf.dl.operators.common.masked_hit_miss_counts(pred, gt, mask, thresholds)¶

Parameters

pred (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)
gt (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)
mask (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)
thresholds (list) –

Returns

hits (mx.nd.NDArray) – Shape: (seq_len, batch_size, len(thresholds))
misses (mx.nd.NDArray) – Shape: (seq_len, batch_size, len(thresholds))
false_alarms (mx.nd.NDArray) – Shape: (seq_len, batch_size, len(thresholds))
correct_negatives (mx.nd.NDArray) – Shape: (seq_len, batch_size, len(thresholds))

swirlspy.qpf.dl.operators.common.multi_segment_slice_axis(data, axis, segment_lengths)¶

Split the data to multiple segments

Parameters

data (mx.sym.Symbol) –
axis (int) –
segment_lengths (list or tuple) – Get the segment_lengths

Returns

ret

Return type

list

swirlspy.qpf.dl.operators.common.normalize_grid(un_norm_grid, width, height)¶

Normalize the grid to [-1, 1]

Parameters

un_norm_grid (mx.sym.Symbol) – Shape : (batch_size, 2, height, width)
width (int) –
height (int) –

Returns

ret

Return type

mx.sym.Symbol

swirlspy.qpf.dl.operators.common.one_step_diff(dat, axis)¶

Parameters

dat (mx.sym.Symbol) –
axes (tuple) –

swirlspy.qpf.dl.operators.common.save_npy(data, save_name='op', save_dir='.')¶

swirlspy.qpf.dl.operators.common.weighted_fss(pred, gt, weight, fss_kernel_HW, batch_size)¶

Parameters

pred (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)
gt (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)
fss_kernel_HW (tuple) – Shape: (1, 1)
weight (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)
batch_size (int) –

Returns

loss_fss – Shape: (seq_len, batch_size)

Return type

mx.sym.Symbol

swirlspy.qpf.dl.operators.common.weighted_l1(pred, gt, weight)¶

swirlspy.qpf.dl.operators.common.weighted_l2(pred, gt, weight)¶

Parameters

pred (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)
gt (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)
weight (mx.sym.Symbol) – Shape: (seq_len, batch_size, 1, H, W)

Returns

l2 – Shape: (seq_len, batch_size)

Return type

mx.nd.NDArray

swirlspy.qpf.dl.operators.common.weighted_mae(pred, gt, weight)¶

swirlspy.qpf.dl.operators.common.weighted_mse(pred, gt, weight)¶

swirlspy.qpf.dl.operators.common.weighted_physical_loss(pred, source, motion_vector_field, step_sizes, weight, batch_size)¶

Parameters

pred (mx.sym.Symbol) – Shape : (seq_len, batch_size, 1, H, W)
source (mx.sym.Symbol : source term in conservation law) – Shape : (seq_len, batch_size, 1, H, W)
motion_vector_field (list of mx.sym.Symbol : velocity fields [u, v]) – Shape : [(seq_len, batch_size, 1, H, W), (seq_len, bach_size, 1, H, W)]
step_sizes (list of float : step sizes [dx, dy, dt]) –
weight (mx.sym.Symbol) – Shape : (seq_len, batch_size, 1, H, W)
batch_size (int) –

Returns

loss – Shape: (seq_len, batch_size)

Return type

mx.sym.Symbol

swirlspy.qpf.dl.operators.conv_rnn module¶

class swirlspy.qpf.dl.operators.conv_rnn.BaseConvRNN(num_filter, b_h_w, h2h_kernel=(3, 3), h2h_dilate=(1, 1), i2h_kernel=(3, 3), i2h_stride=(1, 1), i2h_pad=(1, 1), i2h_dilate=(1, 1), act_type='tanh', prefix='ConvRNN', params=None)¶: Bases: swirlspy.qpf.dl.operators.base_rnn.MyBaseRNNCell

class swirlspy.qpf.dl.operators.conv_rnn.ConvGRU(num_filter, b_h_w, zoneout=0.0, h2h_kernel=(3, 3), h2h_dilate=(1, 1), i2h_kernel=(3, 3), i2h_stride=(1, 1), i2h_pad=(1, 1), i2h_dilate=(1, 1), i2h_adj=(0, 0), no_i2h_bias=False, use_deconv=False, act_type='leaky', prefix='ConvGRU', lr_mult=1.0)¶

Bases: swirlspy.qpf.dl.operators.conv_rnn.BaseConvRNN

property state_info¶: shape and layout information of states

property state_postfix¶

class swirlspy.qpf.dl.operators.conv_rnn.ConvRNN(num_filter, b_h_w, h2h_kernel=(3, 3), h2h_dilate=(1, 1), i2h_kernel=(3, 3), i2h_stride=(1, 1), i2h_pad=(1, 1), i2h_dilate=(1, 1), act_type='leaky', layer_norm=False, prefix='ConvRNN', params=None)¶

Bases: swirlspy.qpf.dl.operators.conv_rnn.BaseConvRNN

property state_info¶: shape and layout information of states

swirlspy.qpf.dl.operators.traj_rnn module¶

class swirlspy.qpf.dl.operators.traj_rnn.TrajGRU(b_h_w, num_filter, zoneout=0.0, L=5, i2h_kernel=(3, 3), i2h_stride=(1, 1), i2h_pad=(1, 1), h2h_kernel=(5, 5), h2h_dilate=(1, 1), act_type='leaky', prefix='TrajGRU', lr_mult=1.0)¶

Bases: swirlspy.qpf.dl.operators.conv_rnn.BaseConvRNN

property state_info¶: shape and layout information of states

property state_postfix¶

swirlspy.qpf.dl.operators.traj_rnn.flow_conv(data, num_filter, flows, weight, bias, name)¶

swirlspy.qpf.dl.operators.transformations module¶

swirlspy.qpf.dl.operators.transformations.CDNA(data, kernels, mask, batch_size, num_filter, kernel_size)¶

We assume that the kernels and masks are the output of an identity activation

Parameters

data (mx.sym.symbol) – Shape: (batch_size, C, H, W)
kernels (mx.sym.symbol) – Shape: (batch_size, M, K, K)
mask (mx.sym.symbol) – Shape: (batch_size, M, H, W)
batch_size (int) –
num_filter (int) – M
kernel_size (int) – K

Returns

ret – Shape: (batch_size, C, H, W)

Return type

mx.sym.symbol

swirlspy.qpf.dl.operators.transformations.DFN(data, local_kernels, K, batch_size)¶

[NIPS2016] Dynamic Filter Network

Parameters

data (mx.sym.symbol) – Shape: (batch_size, C, H, W)
local_kernels (mx.sym.symbol) – Shape: (batch_size, K*K, H, W)
K (int) – size of the local convolutional kernel
batch_size (int) – size of the minibatch

swirlspy.qpf.dl.operators.transformations.STP(data, affine_transform_matrices, mask, num_filter, kernel_size)¶

Spatial Transformer Predictor

Parameters

data (mx.sym.symbol) –
affine_transform_matrices –
mask –