mmagic.models.editors.basicvsr.basicvsr_net

Module Contents

Classes

BasicVSRNet	BasicVSR network structure for video super-resolution.
ResidualBlocksWithInputConv	Residual blocks with a convolution in front.
SPyNet	SPyNet network structure.
SPyNetBasicModule	Basic Module for SPyNet.

class mmagic.models.editors.basicvsr.basicvsr_net.BasicVSRNet(mid_channels=64, num_blocks=30, spynet_pretrained=None)

Bases: mmengine.model.BaseModule

BasicVSR network structure for video super-resolution.

Support only x4 upsampling.

Paper:

BasicVSR: The Search for Essential Components in Video Super-Resolution and Beyond, CVPR, 2021

Parameters

• mid_channels (int) – Channel number of the intermediate features. Default: 64.

• num_blocks (int) – Number of residual blocks in each propagation branch. Default: 30.

• spynet_pretrained (str) – Pre-trained model path of SPyNet. Default: None.

check_if_mirror_extended(lrs)

Check whether the input is a mirror-extended sequence.

If mirror-extended, the i-th (i=0, …, t-1) frame is equal to the (t-1-i)-th frame.

Parameters

lrs (tensor) – Input LR images with shape (n, t, c, h, w)

compute_flow(lrs)

Compute optical flow using SPyNet for feature warping.

Note that if the input is an mirror-extended sequence, ‘flows_forward’ is not needed, since it is equal to ‘flows_backward.flip(1)’.

Parameters

lrs (tensor) – Input LR images with shape (n, t, c, h, w)

Returns

Optical flow. ‘flows_forward’ corresponds to the

flows used for forward-time propagation (current to previous). ‘flows_backward’ corresponds to the flows used for backward-time propagation (current to next).

Return type

tuple(Tensor)

forward(lrs)

Forward function for BasicVSR.

Parameters

lrs (Tensor) – Input LR sequence with shape (n, t, c, h, w).

Returns

Output HR sequence with shape (n, t, c, 4h, 4w).

Return type

Tensor

class mmagic.models.editors.basicvsr.basicvsr_net.ResidualBlocksWithInputConv(in_channels, out_channels=64, num_blocks=30)

Bases: mmengine.model.BaseModule

Residual blocks with a convolution in front.

Parameters

• in_channels (int) – Number of input channels of the first conv.

• out_channels (int) – Number of channels of the residual blocks. Default: 64.

• num_blocks (int) – Number of residual blocks. Default: 30.

forward(feat)

Forward function for ResidualBlocksWithInputConv.

Parameters

feat (Tensor) – Input feature with shape (n, in_channels, h, w)

Returns

Output feature with shape (n, out_channels, h, w)

Return type

Tensor

class mmagic.models.editors.basicvsr.basicvsr_net.SPyNet(pretrained)

Bases: mmengine.model.BaseModule

SPyNet network structure.

The difference to the SPyNet in [tof.py] is that

1. more SPyNetBasicModule is used in this version, and

2. no batch normalization is used in this version.

Paper:

Optical Flow Estimation using a Spatial Pyramid Network, CVPR, 2017

Parameters

pretrained (str) – path for pre-trained SPyNet. Default: None.

compute_flow(ref, supp)

Compute flow from ref to supp.

Note that in this function, the images are already resized to a multiple of 32.

Parameters

• ref (Tensor) – Reference image with shape of (n, 3, h, w).

• supp (Tensor) – Supporting image with shape of (n, 3, h, w).

Returns

Estimated optical flow: (n, 2, h, w).

Return type

Tensor

forward(ref, supp)

Forward function of SPyNet.

This function computes the optical flow from ref to supp.

Parameters

• ref (Tensor) – Reference image with shape of (n, 3, h, w).

• supp (Tensor) – Supporting image with shape of (n, 3, h, w).

Returns

Estimated optical flow: (n, 2, h, w).

Return type

Tensor

class mmagic.models.editors.basicvsr.basicvsr_net.SPyNetBasicModule

Bases: mmengine.model.BaseModule

Basic Module for SPyNet.

Paper:

Optical Flow Estimation using a Spatial Pyramid Network, CVPR, 2017

forward(tensor_input)

Parameters

tensor_input (Tensor) – Input tensor with shape (b, 8, h, w). 8 channels contain: [reference image (3), neighbor image (3), initial flow (2)].

Returns

Refined flow with shape (b, 2, h, w)

Return type

Tensor