diff --git a/LICENSE/LICENSE_SwinIR b/LICENSE/LICENSE_SwinIR new file mode 100644 index 000000000..e5e4ee061 --- /dev/null +++ b/LICENSE/LICENSE_SwinIR @@ -0,0 +1,201 @@ + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. + + END OF TERMS AND CONDITIONS + + APPENDIX: How to apply the Apache License to your work. + + To apply the Apache License to your work, attach the following + boilerplate notice, with the fields enclosed by brackets "[]" + replaced with your own identifying information. (Don't include + the brackets!) The text should be enclosed in the appropriate + comment syntax for the file format. We also recommend that a + file or class name and description of purpose be included on the + same "printed page" as the copyright notice for easier + identification within third-party archives. + + Copyright [2021] [SwinIR Authors] + + Licensed under the Apache License, Version 2.0 (the "License"); + you may not use this file except in compliance with the License. + You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. diff --git a/LICENSE/LICENSE_pytorch-image-models b/LICENSE/LICENSE_pytorch-image-models new file mode 100644 index 000000000..433681909 --- /dev/null +++ b/LICENSE/LICENSE_pytorch-image-models @@ -0,0 +1,201 @@ + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. + + END OF TERMS AND CONDITIONS + + APPENDIX: How to apply the Apache License to your work. + + To apply the Apache License to your work, attach the following + boilerplate notice, with the fields enclosed by brackets "{}" + replaced with your own identifying information. (Don't include + the brackets!) The text should be enclosed in the appropriate + comment syntax for the file format. We also recommend that a + file or class name and description of purpose be included on the + same "printed page" as the copyright notice for easier + identification within third-party archives. + + Copyright 2019 Ross Wightman + + Licensed under the Apache License, Version 2.0 (the "License"); + you may not use this file except in compliance with the License. + You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. diff --git a/LICENSE/README.md b/LICENSE/README.md index 15e0d345d..93179cc13 100644 --- a/LICENSE/README.md +++ b/LICENSE/README.md @@ -9,6 +9,10 @@ This BasicSR project is released under the Apache 2.0 license. - The codes are largely modified from the repository [DFDNet](https://github.com/csxmli2016/DFDNet). Their license is [Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License](https://creativecommons.org/licenses/by-nc-sa/4.0/). - DiffJPEG - Modified from https://github.com/mlomnitz/DiffJPEG. +- [pytorch-image-models](https://github.com/rwightman/pytorch-image-models/) + - We use the implementation of `DropPath` and `trunc_normal_` from [pytorch-image-models](https://github.com/rwightman/pytorch-image-models/). The LICENSE is included as [LICENSE_pytorch-image-models](LICENSE/LICENSE_pytorch-image-models). +- [SwinIR](https://github.com/JingyunLiang/SwinIR) + - The arch implementation of SwinIR is from [SwinIR](https://github.com/JingyunLiang/SwinIR). The LICENSE is included as [LICENSE_SwinIR](LICENSE/LICENSE_SwinIR). ## References diff --git a/README.md b/README.md index 85351a98c..5258acfab 100644 --- a/README.md +++ b/README.md @@ -25,6 +25,7 @@ BasicSR (**Basic** **S**uper **R**estoration) is an open-source **image and vide :triangular_flag_on_post: **New Features/Updates** +- :white_check_mark: Sep 2, 2021. Add **SwinIR training and testing** codes: [SwinIR](https://github.com/JingyunLiang/SwinIR) by [Jingyun Liang](https://github.com/JingyunLiang):+1:. More details are in [HOWTOs.md](docs/HOWTOs.md#how-to-train-swinir-sr) - :white_check_mark: Aug 5, 2021. Add NIQE, which produces the same results as MATLAB (both are 5.7296 for tests/data/baboon.png). - :white_check_mark: July 31, 2021. Add **bi-directional video super-resolution** codes: [**BasicVSR** and IconVSR](https://arxiv.org/abs/2012.02181). - :white_check_mark: July 20, 2021. Add **dual-blind face restoration** codes: [HiFaceGAN](https://github.com/Lotayou/Face-Renovation) codes by [Lotayou](https://lotayou.github.io/). @@ -73,7 +74,7 @@ These pipelines/commands cannot cover all the cases and more details are in the | **Super Resolution** | | | | | | | ESRGAN | *TODO* | *TODO* | SRGAN | *TODO* | *TODO*| | EDSR | *TODO* | *TODO* | SRResNet | *TODO* | *TODO*| -| RCAN | *TODO* | *TODO* | | | | +| RCAN | *TODO* | *TODO* | SwinIR | [Train](docs/HOWTOs.md#how-to-train-swinir-sr) | [Inference](docs/HOWTOs.md#how-to-inference-swinir-sr)| | EDVR | *TODO* | *TODO* | DUF | - | *TODO* | | BasicVSR | *TODO* | *TODO* | TOF | - | *TODO* | | **Deblurring** | | | | | | diff --git a/README_CN.md b/README_CN.md index 68670c4d9..3c94495ef 100644 --- a/README_CN.md +++ b/README_CN.md @@ -25,6 +25,7 @@ BasicSR (**Basic** **S**uper **R**estoration) 是一个基于 PyTorch 的开源 :triangular_flag_on_post: **新的特性/更新** +- :white_check_mark: Sep 2, 2021. 添加 **SwinIR 训练和测试** 代码: [SwinIR](https://github.com/JingyunLiang/SwinIR) by [Jingyun Liang](https://github.com/JingyunLiang):+1:. 更多内容参见 [HOWTOs.md](docs/HOWTOs.md#how-to-train-swinir-sr) - :white_check_mark: Aug 5, 2021. 添加了NIQE， 它输出和MATLAB一样的结果 (both are 5.7296 for tests/data/baboon.png). - :white_check_mark: July 31, 2021. Add **bi-directional video super-resolution** codes: [**BasicVSR** and IconVSR](https://arxiv.org/abs/2012.02181). - :white_check_mark: July 20, 2021. Add **dual-blind face restoration** codes: [**HiFaceGAN**](https://github.com/Lotayou/Face-Renovation) codes by [Lotayou](https://lotayou.github.io/). @@ -72,7 +73,7 @@ BasicSR (**Basic** **S**uper **R**estoration) 是一个基于 PyTorch 的开源 | **Super Resolution** | | | | | | | ESRGAN | *TODO* | *TODO* | SRGAN | *TODO* | *TODO*| | EDSR | *TODO* | *TODO* | SRResNet | *TODO* | *TODO*| -| RCAN | *TODO* | *TODO* | | | | +| RCAN | *TODO* | *TODO* | SwinIR | [Train](docs/HOWTOs.md#how-to-train-swinir-sr) | [Inference](docs/HOWTOs.md#how-to-inference-swinir-sr)| | EDVR | *TODO* | *TODO* | DUF | - | *TODO* | | BasicVSR | *TODO* | *TODO* | TOF | - | *TODO* | | **Deblurring** | | | | | | diff --git a/VERSION b/VERSION index 6b3612faf..984213ca0 100644 --- a/VERSION +++ b/VERSION @@ -1 +1 @@ -1.3.4.0 +1.3.4.1 diff --git a/basicsr/archs/arch_util.py b/basicsr/archs/arch_util.py index 7cf00720b..d23b51a5a 100644 --- a/basicsr/archs/arch_util.py +++ b/basicsr/archs/arch_util.py @@ -1,5 +1,8 @@ +import collections.abc import math import torch +import warnings +from itertools import repeat from torch import nn as nn from torch.nn import functional as F from torch.nn import init as init @@ -225,3 +228,85 @@ def forward(self, x, feat): return modulated_deform_conv(x, offset, mask, self.weight, self.bias, self.stride, self.padding, self.dilation, self.groups, self.deformable_groups) + + +def _no_grad_trunc_normal_(tensor, mean, std, a, b): + # From: https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/layers/weight_init.py + # Cut & paste from PyTorch official master until it's in a few official releases - RW + # Method based on https://people.sc.fsu.edu/~jburkardt/presentations/truncated_normal.pdf + def norm_cdf(x): + # Computes standard normal cumulative distribution function + return (1. + math.erf(x / math.sqrt(2.))) / 2. + + if (mean < a - 2 * std) or (mean > b + 2 * std): + warnings.warn( + 'mean is more than 2 std from [a, b] in nn.init.trunc_normal_. ' + 'The distribution of values may be incorrect.', + stacklevel=2) + + with torch.no_grad(): + # Values are generated by using a truncated uniform distribution and + # then using the inverse CDF for the normal distribution. + # Get upper and lower cdf values + low = norm_cdf((a - mean) / std) + up = norm_cdf((b - mean) / std) + + # Uniformly fill tensor with values from [low, up], then translate to + # [2l-1, 2u-1]. + tensor.uniform_(2 * low - 1, 2 * up - 1) + + # Use inverse cdf transform for normal distribution to get truncated + # standard normal + tensor.erfinv_() + + # Transform to proper mean, std + tensor.mul_(std * math.sqrt(2.)) + tensor.add_(mean) + + # Clamp to ensure it's in the proper range + tensor.clamp_(min=a, max=b) + return tensor + + +def trunc_normal_(tensor, mean=0., std=1., a=-2., b=2.): + r"""Fills the input Tensor with values drawn from a truncated + normal distribution. + + From: https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/layers/weight_init.py + + The values are effectively drawn from the + normal distribution :math:`\mathcal{N}(\text{mean}, \text{std}^2)` + with values outside :math:`[a, b]` redrawn until they are within + the bounds. The method used for generating the random values works + best when :math:`a \leq \text{mean} \leq b`. + + Args: + tensor: an n-dimensional `torch.Tensor` + mean: the mean of the normal distribution + std: the standard deviation of the normal distribution + a: the minimum cutoff value + b: the maximum cutoff value + + Examples: + >>> w = torch.empty(3, 5) + >>> nn.init.trunc_normal_(w) + """ + return _no_grad_trunc_normal_(tensor, mean, std, a, b) + + +# From PyTorch +def _ntuple(n): + + def parse(x): + if isinstance(x, collections.abc.Iterable): + return x + return tuple(repeat(x, n)) + + return parse + + +to_1tuple = _ntuple(1) +to_2tuple = _ntuple(2) +to_3tuple = _ntuple(3) +to_4tuple = _ntuple(4) +to_ntuple = _ntuple diff --git a/basicsr/archs/swinir_arch.py b/basicsr/archs/swinir_arch.py new file mode 100644 index 000000000..58bc13a26 --- /dev/null +++ b/basicsr/archs/swinir_arch.py @@ -0,0 +1,957 @@ +# Modified from https://github.com/JingyunLiang/SwinIR +# SwinIR: Image Restoration Using Swin Transformer, https://arxiv.org/abs/2108.10257 +# Originally Written by Ze Liu, Modified by Jingyun Liang. + +import math +import torch +import torch.nn as nn +import torch.utils.checkpoint as checkpoint + +from basicsr.utils.registry import ARCH_REGISTRY +from .arch_util import to_2tuple, trunc_normal_ + + +def drop_path(x, drop_prob: float = 0., training: bool = False): + """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). + + From: https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/layers/drop.py + """ + if drop_prob == 0. or not training: + return x + keep_prob = 1 - drop_prob + shape = (x.shape[0], ) + (1, ) * (x.ndim - 1) # work with diff dim tensors, not just 2D ConvNets + random_tensor = keep_prob + torch.rand(shape, dtype=x.dtype, device=x.device) + random_tensor.floor_() # binarize + output = x.div(keep_prob) * random_tensor + return output + + +class DropPath(nn.Module): + """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). + + From: https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/layers/drop.py + """ + + def __init__(self, drop_prob=None): + super(DropPath, self).__init__() + self.drop_prob = drop_prob + + def forward(self, x): + return drop_path(x, self.drop_prob, self.training) + + +class Mlp(nn.Module): + + def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.): + super().__init__() + out_features = out_features or in_features + hidden_features = hidden_features or in_features + self.fc1 = nn.Linear(in_features, hidden_features) + self.act = act_layer() + self.fc2 = nn.Linear(hidden_features, out_features) + self.drop = nn.Dropout(drop) + + def forward(self, x): + x = self.fc1(x) + x = self.act(x) + x = self.drop(x) + x = self.fc2(x) + x = self.drop(x) + return x + + +def window_partition(x, window_size): + """ + Args: + x: (B, H, W, C) + window_size (int): window size + + Returns: + windows: (num_windows*B, window_size, window_size, C) + """ + B, H, W, C = x.shape + x = x.view(B, H // window_size, window_size, W // window_size, window_size, C) + windows = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, C) + return windows + + +def window_reverse(windows, window_size, H, W): + """ + Args: + windows: (num_windows*B, window_size, window_size, C) + window_size (int): Window size + H (int): Height of image + W (int): Width of image + + Returns: + x: (B, H, W, C) + """ + B = int(windows.shape[0] / (H * W / window_size / window_size)) + x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1) + x = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(B, H, W, -1) + return x + + +class WindowAttention(nn.Module): + r""" Window based multi-head self attention (W-MSA) module with relative position bias. + It supports both of shifted and non-shifted window. + + Args: + dim (int): Number of input channels. + window_size (tuple[int]): The height and width of the window. + num_heads (int): Number of attention heads. + qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True + qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set + attn_drop (float, optional): Dropout ratio of attention weight. Default: 0.0 + proj_drop (float, optional): Dropout ratio of output. Default: 0.0 + """ + + def __init__(self, dim, window_size, num_heads, qkv_bias=True, qk_scale=None, attn_drop=0., proj_drop=0.): + + super().__init__() + self.dim = dim + self.window_size = window_size # Wh, Ww + self.num_heads = num_heads + head_dim = dim // num_heads + self.scale = qk_scale or head_dim**-0.5 + + # define a parameter table of relative position bias + self.relative_position_bias_table = nn.Parameter( + torch.zeros((2 * window_size[0] - 1) * (2 * window_size[1] - 1), num_heads)) # 2*Wh-1 * 2*Ww-1, nH + + # get pair-wise relative position index for each token inside the window + coords_h = torch.arange(self.window_size[0]) + coords_w = torch.arange(self.window_size[1]) + coords = torch.stack(torch.meshgrid([coords_h, coords_w])) # 2, Wh, Ww + coords_flatten = torch.flatten(coords, 1) # 2, Wh*Ww + relative_coords = coords_flatten[:, :, None] - coords_flatten[:, None, :] # 2, Wh*Ww, Wh*Ww + relative_coords = relative_coords.permute(1, 2, 0).contiguous() # Wh*Ww, Wh*Ww, 2 + relative_coords[:, :, 0] += self.window_size[0] - 1 # shift to start from 0 + relative_coords[:, :, 1] += self.window_size[1] - 1 + relative_coords[:, :, 0] *= 2 * self.window_size[1] - 1 + relative_position_index = relative_coords.sum(-1) # Wh*Ww, Wh*Ww + self.register_buffer('relative_position_index', relative_position_index) + + self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias) + self.attn_drop = nn.Dropout(attn_drop) + self.proj = nn.Linear(dim, dim) + + self.proj_drop = nn.Dropout(proj_drop) + + trunc_normal_(self.relative_position_bias_table, std=.02) + self.softmax = nn.Softmax(dim=-1) + + def forward(self, x, mask=None): + """ + Args: + x: input features with shape of (num_windows*B, N, C) + mask: (0/-inf) mask with shape of (num_windows, Wh*Ww, Wh*Ww) or None + """ + B_, N, C = x.shape + qkv = self.qkv(x).reshape(B_, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4) + q, k, v = qkv[0], qkv[1], qkv[2] # make torchscript happy (cannot use tensor as tuple) + + q = q * self.scale + attn = (q @ k.transpose(-2, -1)) + + relative_position_bias = self.relative_position_bias_table[self.relative_position_index.view(-1)].view( + self.window_size[0] * self.window_size[1], self.window_size[0] * self.window_size[1], -1) # Wh*Ww,Wh*Ww,nH + relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous() # nH, Wh*Ww, Wh*Ww + attn = attn + relative_position_bias.unsqueeze(0) + + if mask is not None: + nW = mask.shape[0] + attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(1).unsqueeze(0) + attn = attn.view(-1, self.num_heads, N, N) + attn = self.softmax(attn) + else: + attn = self.softmax(attn) + + attn = self.attn_drop(attn) + + x = (attn @ v).transpose(1, 2).reshape(B_, N, C) + x = self.proj(x) + x = self.proj_drop(x) + return x + + def extra_repr(self) -> str: + return f'dim={self.dim}, window_size={self.window_size}, num_heads={self.num_heads}' + + def flops(self, N): + # calculate flops for 1 window with token length of N + flops = 0 + # qkv = self.qkv(x) + flops += N * self.dim * 3 * self.dim + # attn = (q @ k.transpose(-2, -1)) + flops += self.num_heads * N * (self.dim // self.num_heads) * N + # x = (attn @ v) + flops += self.num_heads * N * N * (self.dim // self.num_heads) + # x = self.proj(x) + flops += N * self.dim * self.dim + return flops + + +class SwinTransformerBlock(nn.Module): + r""" Swin Transformer Block. + + Args: + dim (int): Number of input channels. + input_resolution (tuple[int]): Input resolution. + num_heads (int): Number of attention heads. + window_size (int): Window size. + shift_size (int): Shift size for SW-MSA. + mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. + qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True + qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set. + drop (float, optional): Dropout rate. Default: 0.0 + attn_drop (float, optional): Attention dropout rate. Default: 0.0 + drop_path (float, optional): Stochastic depth rate. Default: 0.0 + act_layer (nn.Module, optional): Activation layer. Default: nn.GELU + norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm + """ + + def __init__(self, + dim, + input_resolution, + num_heads, + window_size=7, + shift_size=0, + mlp_ratio=4., + qkv_bias=True, + qk_scale=None, + drop=0., + attn_drop=0., + drop_path=0., + act_layer=nn.GELU, + norm_layer=nn.LayerNorm): + super().__init__() + self.dim = dim + self.input_resolution = input_resolution + self.num_heads = num_heads + self.window_size = window_size + self.shift_size = shift_size + self.mlp_ratio = mlp_ratio + if min(self.input_resolution) <= self.window_size: + # if window size is larger than input resolution, we don't partition windows + self.shift_size = 0 + self.window_size = min(self.input_resolution) + assert 0 <= self.shift_size < self.window_size, 'shift_size must in 0-window_size' + + self.norm1 = norm_layer(dim) + self.attn = WindowAttention( + dim, + window_size=to_2tuple(self.window_size), + num_heads=num_heads, + qkv_bias=qkv_bias, + qk_scale=qk_scale, + attn_drop=attn_drop, + proj_drop=drop) + + self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity() + self.norm2 = norm_layer(dim) + mlp_hidden_dim = int(dim * mlp_ratio) + self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop) + + if self.shift_size > 0: + attn_mask = self.calculate_mask(self.input_resolution) + else: + attn_mask = None + + self.register_buffer('attn_mask', attn_mask) + + def calculate_mask(self, x_size): + # calculate attention mask for SW-MSA + H, W = x_size + img_mask = torch.zeros((1, H, W, 1)) # 1 H W 1 + h_slices = (slice(0, -self.window_size), slice(-self.window_size, + -self.shift_size), slice(-self.shift_size, None)) + w_slices = (slice(0, -self.window_size), slice(-self.window_size, + -self.shift_size), slice(-self.shift_size, None)) + cnt = 0 + for h in h_slices: + for w in w_slices: + img_mask[:, h, w, :] = cnt + cnt += 1 + + mask_windows = window_partition(img_mask, self.window_size) # nW, window_size, window_size, 1 + mask_windows = mask_windows.view(-1, self.window_size * self.window_size) + attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2) + attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0)) + + return attn_mask + + def forward(self, x, x_size): + H, W = x_size + B, L, C = x.shape + # assert L == H * W, "input feature has wrong size" + + shortcut = x + x = self.norm1(x) + x = x.view(B, H, W, C) + + # cyclic shift + if self.shift_size > 0: + shifted_x = torch.roll(x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2)) + else: + shifted_x = x + + # partition windows + x_windows = window_partition(shifted_x, self.window_size) # nW*B, window_size, window_size, C + x_windows = x_windows.view(-1, self.window_size * self.window_size, C) # nW*B, window_size*window_size, C + + # W-MSA/SW-MSA (to be compatible for testing on images whose shapes are the multiple of window size + if self.input_resolution == x_size: + attn_windows = self.attn(x_windows, mask=self.attn_mask) # nW*B, window_size*window_size, C + else: + attn_windows = self.attn(x_windows, mask=self.calculate_mask(x_size).to(x.device)) + + # merge windows + attn_windows = attn_windows.view(-1, self.window_size, self.window_size, C) + shifted_x = window_reverse(attn_windows, self.window_size, H, W) # B H' W' C + + # reverse cyclic shift + if self.shift_size > 0: + x = torch.roll(shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2)) + else: + x = shifted_x + x = x.view(B, H * W, C) + + # FFN + x = shortcut + self.drop_path(x) + x = x + self.drop_path(self.mlp(self.norm2(x))) + + return x + + def extra_repr(self) -> str: + return f'dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, ' \ + f'window_size={self.window_size}, shift_size={self.shift_size}, mlp_ratio={self.mlp_ratio}' + + def flops(self): + flops = 0 + H, W = self.input_resolution + # norm1 + flops += self.dim * H * W + # W-MSA/SW-MSA + nW = H * W / self.window_size / self.window_size + flops += nW * self.attn.flops(self.window_size * self.window_size) + # mlp + flops += 2 * H * W * self.dim * self.dim * self.mlp_ratio + # norm2 + flops += self.dim * H * W + return flops + + +class PatchMerging(nn.Module): + r""" Patch Merging Layer. + + Args: + input_resolution (tuple[int]): Resolution of input feature. + dim (int): Number of input channels. + norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm + """ + + def __init__(self, input_resolution, dim, norm_layer=nn.LayerNorm): + super().__init__() + self.input_resolution = input_resolution + self.dim = dim + self.reduction = nn.Linear(4 * dim, 2 * dim, bias=False) + self.norm = norm_layer(4 * dim) + + def forward(self, x): + """ + x: B, H*W, C + """ + H, W = self.input_resolution + B, L, C = x.shape + assert L == H * W, 'input feature has wrong size' + assert H % 2 == 0 and W % 2 == 0, f'x size ({H}*{W}) are not even.' + + x = x.view(B, H, W, C) + + x0 = x[:, 0::2, 0::2, :] # B H/2 W/2 C + x1 = x[:, 1::2, 0::2, :] # B H/2 W/2 C + x2 = x[:, 0::2, 1::2, :] # B H/2 W/2 C + x3 = x[:, 1::2, 1::2, :] # B H/2 W/2 C + x = torch.cat([x0, x1, x2, x3], -1) # B H/2 W/2 4*C + x = x.view(B, -1, 4 * C) # B H/2*W/2 4*C + + x = self.norm(x) + x = self.reduction(x) + + return x + + def extra_repr(self) -> str: + return f'input_resolution={self.input_resolution}, dim={self.dim}' + + def flops(self): + H, W = self.input_resolution + flops = H * W * self.dim + flops += (H // 2) * (W // 2) * 4 * self.dim * 2 * self.dim + return flops + + +class BasicLayer(nn.Module): + """ A basic Swin Transformer layer for one stage. + + Args: + dim (int): Number of input channels. + input_resolution (tuple[int]): Input resolution. + depth (int): Number of blocks. + num_heads (int): Number of attention heads. + window_size (int): Local window size. + mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. + qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True + qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set. + drop (float, optional): Dropout rate. Default: 0.0 + attn_drop (float, optional): Attention dropout rate. Default: 0.0 + drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0 + norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm + downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None + use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False. + """ + + def __init__(self, + dim, + input_resolution, + depth, + num_heads, + window_size, + mlp_ratio=4., + qkv_bias=True, + qk_scale=None, + drop=0., + attn_drop=0., + drop_path=0., + norm_layer=nn.LayerNorm, + downsample=None, + use_checkpoint=False): + + super().__init__() + self.dim = dim + self.input_resolution = input_resolution + self.depth = depth + self.use_checkpoint = use_checkpoint + + # build blocks + self.blocks = nn.ModuleList([ + SwinTransformerBlock( + dim=dim, + input_resolution=input_resolution, + num_heads=num_heads, + window_size=window_size, + shift_size=0 if (i % 2 == 0) else window_size // 2, + mlp_ratio=mlp_ratio, + qkv_bias=qkv_bias, + qk_scale=qk_scale, + drop=drop, + attn_drop=attn_drop, + drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path, + norm_layer=norm_layer) for i in range(depth) + ]) + + # patch merging layer + if downsample is not None: + self.downsample = downsample(input_resolution, dim=dim, norm_layer=norm_layer) + else: + self.downsample = None + + def forward(self, x, x_size): + for blk in self.blocks: + if self.use_checkpoint: + x = checkpoint.checkpoint(blk, x) + else: + x = blk(x, x_size) + if self.downsample is not None: + x = self.downsample(x) + return x + + def extra_repr(self) -> str: + return f'dim={self.dim}, input_resolution={self.input_resolution}, depth={self.depth}' + + def flops(self): + flops = 0 + for blk in self.blocks: + flops += blk.flops() + if self.downsample is not None: + flops += self.downsample.flops() + return flops + + +class RSTB(nn.Module): + """Residual Swin Transformer Block (RSTB). + + Args: + dim (int): Number of input channels. + input_resolution (tuple[int]): Input resolution. + depth (int): Number of blocks. + num_heads (int): Number of attention heads. + window_size (int): Local window size. + mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. + qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True + qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set. + drop (float, optional): Dropout rate. Default: 0.0 + attn_drop (float, optional): Attention dropout rate. Default: 0.0 + drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0 + norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm + downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None + use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False. + img_size: Input image size. + patch_size: Patch size. + resi_connection: The convolutional block before residual connection. + """ + + def __init__(self, + dim, + input_resolution, + depth, + num_heads, + window_size, + mlp_ratio=4., + qkv_bias=True, + qk_scale=None, + drop=0., + attn_drop=0., + drop_path=0., + norm_layer=nn.LayerNorm, + downsample=None, + use_checkpoint=False, + img_size=224, + patch_size=4, + resi_connection='1conv'): + super(RSTB, self).__init__() + + self.dim = dim + self.input_resolution = input_resolution + + self.residual_group = BasicLayer( + dim=dim, + input_resolution=input_resolution, + depth=depth, + num_heads=num_heads, + window_size=window_size, + mlp_ratio=mlp_ratio, + qkv_bias=qkv_bias, + qk_scale=qk_scale, + drop=drop, + attn_drop=attn_drop, + drop_path=drop_path, + norm_layer=norm_layer, + downsample=downsample, + use_checkpoint=use_checkpoint) + + if resi_connection == '1conv': + self.conv = nn.Conv2d(dim, dim, 3, 1, 1) + elif resi_connection == '3conv': + # to save parameters and memory + self.conv = nn.Sequential( + nn.Conv2d(dim, dim // 4, 3, 1, 1), nn.LeakyReLU(negative_slope=0.2, inplace=True), + nn.Conv2d(dim // 4, dim // 4, 1, 1, 0), nn.LeakyReLU(negative_slope=0.2, inplace=True), + nn.Conv2d(dim // 4, dim, 3, 1, 1)) + + self.patch_embed = PatchEmbed( + img_size=img_size, patch_size=patch_size, in_chans=0, embed_dim=dim, norm_layer=None) + + self.patch_unembed = PatchUnEmbed( + img_size=img_size, patch_size=patch_size, in_chans=0, embed_dim=dim, norm_layer=None) + + def forward(self, x, x_size): + return self.patch_embed(self.conv(self.patch_unembed(self.residual_group(x, x_size), x_size))) + x + + def flops(self): + flops = 0 + flops += self.residual_group.flops() + H, W = self.input_resolution + flops += H * W * self.dim * self.dim * 9 + flops += self.patch_embed.flops() + flops += self.patch_unembed.flops() + + return flops + + +class PatchEmbed(nn.Module): + r""" Image to Patch Embedding + + Args: + img_size (int): Image size. Default: 224. + patch_size (int): Patch token size. Default: 4. + in_chans (int): Number of input image channels. Default: 3. + embed_dim (int): Number of linear projection output channels. Default: 96. + norm_layer (nn.Module, optional): Normalization layer. Default: None + """ + + def __init__(self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None): + super().__init__() + img_size = to_2tuple(img_size) + patch_size = to_2tuple(patch_size) + patches_resolution = [img_size[0] // patch_size[0], img_size[1] // patch_size[1]] + self.img_size = img_size + self.patch_size = patch_size + self.patches_resolution = patches_resolution + self.num_patches = patches_resolution[0] * patches_resolution[1] + + self.in_chans = in_chans + self.embed_dim = embed_dim + + if norm_layer is not None: + self.norm = norm_layer(embed_dim) + else: + self.norm = None + + def forward(self, x): + x = x.flatten(2).transpose(1, 2) # B Ph*Pw C + if self.norm is not None: + x = self.norm(x) + return x + + def flops(self): + flops = 0 + H, W = self.img_size + if self.norm is not None: + flops += H * W * self.embed_dim + return flops + + +class PatchUnEmbed(nn.Module): + r""" Image to Patch Unembedding + + Args: + img_size (int): Image size. Default: 224. + patch_size (int): Patch token size. Default: 4. + in_chans (int): Number of input image channels. Default: 3. + embed_dim (int): Number of linear projection output channels. Default: 96. + norm_layer (nn.Module, optional): Normalization layer. Default: None + """ + + def __init__(self, img_size=224, patch_size=4, in_chans=3, embed_dim=96, norm_layer=None): + super().__init__() + img_size = to_2tuple(img_size) + patch_size = to_2tuple(patch_size) + patches_resolution = [img_size[0] // patch_size[0], img_size[1] // patch_size[1]] + self.img_size = img_size + self.patch_size = patch_size + self.patches_resolution = patches_resolution + self.num_patches = patches_resolution[0] * patches_resolution[1] + + self.in_chans = in_chans + self.embed_dim = embed_dim + + def forward(self, x, x_size): + B, HW, C = x.shape + x = x.transpose(1, 2).view(B, self.embed_dim, x_size[0], x_size[1]) # B Ph*Pw C + return x + + def flops(self): + flops = 0 + return flops + + +class Upsample(nn.Sequential): + """Upsample module. + + Args: + scale (int): Scale factor. Supported scales: 2^n and 3. + num_feat (int): Channel number of intermediate features. + """ + + def __init__(self, scale, num_feat): + m = [] + if (scale & (scale - 1)) == 0: # scale = 2^n + for _ in range(int(math.log(scale, 2))): + m.append(nn.Conv2d(num_feat, 4 * num_feat, 3, 1, 1)) + m.append(nn.PixelShuffle(2)) + elif scale == 3: + m.append(nn.Conv2d(num_feat, 9 * num_feat, 3, 1, 1)) + m.append(nn.PixelShuffle(3)) + else: + raise ValueError(f'scale {scale} is not supported. ' 'Supported scales: 2^n and 3.') + super(Upsample, self).__init__(*m) + + +class UpsampleOneStep(nn.Sequential): + """UpsampleOneStep module (the difference with Upsample is that it always only has 1conv + 1pixelshuffle) + Used in lightweight SR to save parameters. + + Args: + scale (int): Scale factor. Supported scales: 2^n and 3. + num_feat (int): Channel number of intermediate features. + + """ + + def __init__(self, scale, num_feat, num_out_ch, input_resolution=None): + self.num_feat = num_feat + self.input_resolution = input_resolution + m = [] + m.append(nn.Conv2d(num_feat, (scale**2) * num_out_ch, 3, 1, 1)) + m.append(nn.PixelShuffle(scale)) + super(UpsampleOneStep, self).__init__(*m) + + def flops(self): + H, W = self.input_resolution + flops = H * W * self.num_feat * 3 * 9 + return flops + + +@ARCH_REGISTRY.register() +class SwinIR(nn.Module): + r""" SwinIR + A PyTorch impl of : `SwinIR: Image Restoration Using Swin Transformer`, based on Swin Transformer. + + Args: + img_size (int | tuple(int)): Input image size. Default 64 + patch_size (int | tuple(int)): Patch size. Default: 1 + in_chans (int): Number of input image channels. Default: 3 + embed_dim (int): Patch embedding dimension. Default: 96 + depths (tuple(int)): Depth of each Swin Transformer layer. + num_heads (tuple(int)): Number of attention heads in different layers. + window_size (int): Window size. Default: 7 + mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4 + qkv_bias (bool): If True, add a learnable bias to query, key, value. Default: True + qk_scale (float): Override default qk scale of head_dim ** -0.5 if set. Default: None + drop_rate (float): Dropout rate. Default: 0 + attn_drop_rate (float): Attention dropout rate. Default: 0 + drop_path_rate (float): Stochastic depth rate. Default: 0.1 + norm_layer (nn.Module): Normalization layer. Default: nn.LayerNorm. + ape (bool): If True, add absolute position embedding to the patch embedding. Default: False + patch_norm (bool): If True, add normalization after patch embedding. Default: True + use_checkpoint (bool): Whether to use checkpointing to save memory. Default: False + upscale: Upscale factor. 2/3/4/8 for image SR, 1 for denoising and compress artifact reduction + img_range: Image range. 1. or 255. + upsampler: The reconstruction reconstruction module. 'pixelshuffle'/'pixelshuffledirect'/'nearest+conv'/None + resi_connection: The convolutional block before residual connection. '1conv'/'3conv' + """ + + def __init__(self, + img_size=64, + patch_size=1, + in_chans=3, + embed_dim=96, + depths=[6, 6, 6, 6], + num_heads=[6, 6, 6, 6], + window_size=7, + mlp_ratio=4., + qkv_bias=True, + qk_scale=None, + drop_rate=0., + attn_drop_rate=0., + drop_path_rate=0.1, + norm_layer=nn.LayerNorm, + ape=False, + patch_norm=True, + use_checkpoint=False, + upscale=2, + img_range=1., + upsampler='', + resi_connection='1conv', + **kwargs): + super(SwinIR, self).__init__() + num_in_ch = in_chans + num_out_ch = in_chans + num_feat = 64 + self.img_range = img_range + if in_chans == 3: + rgb_mean = (0.4488, 0.4371, 0.4040) + self.mean = torch.Tensor(rgb_mean).view(1, 3, 1, 1) + else: + self.mean = torch.zeros(1, 1, 1, 1) + self.upscale = upscale + self.upsampler = upsampler + + # ------------------------- 1, shallow feature extraction ------------------------- # + self.conv_first = nn.Conv2d(num_in_ch, embed_dim, 3, 1, 1) + + # ------------------------- 2, deep feature extraction ------------------------- # + self.num_layers = len(depths) + self.embed_dim = embed_dim + self.ape = ape + self.patch_norm = patch_norm + self.num_features = embed_dim + self.mlp_ratio = mlp_ratio + + # split image into non-overlapping patches + self.patch_embed = PatchEmbed( + img_size=img_size, + patch_size=patch_size, + in_chans=embed_dim, + embed_dim=embed_dim, + norm_layer=norm_layer if self.patch_norm else None) + num_patches = self.patch_embed.num_patches + patches_resolution = self.patch_embed.patches_resolution + self.patches_resolution = patches_resolution + + # merge non-overlapping patches into image + self.patch_unembed = PatchUnEmbed( + img_size=img_size, + patch_size=patch_size, + in_chans=embed_dim, + embed_dim=embed_dim, + norm_layer=norm_layer if self.patch_norm else None) + + # absolute position embedding + if self.ape: + self.absolute_pos_embed = nn.Parameter(torch.zeros(1, num_patches, embed_dim)) + trunc_normal_(self.absolute_pos_embed, std=.02) + + self.pos_drop = nn.Dropout(p=drop_rate) + + # stochastic depth + dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))] # stochastic depth decay rule + + # build Residual Swin Transformer blocks (RSTB) + self.layers = nn.ModuleList() + for i_layer in range(self.num_layers): + layer = RSTB( + dim=embed_dim, + input_resolution=(patches_resolution[0], patches_resolution[1]), + depth=depths[i_layer], + num_heads=num_heads[i_layer], + window_size=window_size, + mlp_ratio=self.mlp_ratio, + qkv_bias=qkv_bias, + qk_scale=qk_scale, + drop=drop_rate, + attn_drop=attn_drop_rate, + drop_path=dpr[sum(depths[:i_layer]):sum(depths[:i_layer + 1])], # no impact on SR results + norm_layer=norm_layer, + downsample=None, + use_checkpoint=use_checkpoint, + img_size=img_size, + patch_size=patch_size, + resi_connection=resi_connection) + self.layers.append(layer) + self.norm = norm_layer(self.num_features) + + # build the last conv layer in deep feature extraction + if resi_connection == '1conv': + self.conv_after_body = nn.Conv2d(embed_dim, embed_dim, 3, 1, 1) + elif resi_connection == '3conv': + # to save parameters and memory + self.conv_after_body = nn.Sequential( + nn.Conv2d(embed_dim, embed_dim // 4, 3, 1, 1), nn.LeakyReLU(negative_slope=0.2, inplace=True), + nn.Conv2d(embed_dim // 4, embed_dim // 4, 1, 1, 0), nn.LeakyReLU(negative_slope=0.2, inplace=True), + nn.Conv2d(embed_dim // 4, embed_dim, 3, 1, 1)) + + # ------------------------- 3, high quality image reconstruction ------------------------- # + if self.upsampler == 'pixelshuffle': + # for classical SR + self.conv_before_upsample = nn.Sequential( + nn.Conv2d(embed_dim, num_feat, 3, 1, 1), nn.LeakyReLU(inplace=True)) + self.upsample = Upsample(upscale, num_feat) + self.conv_last = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1) + elif self.upsampler == 'pixelshuffledirect': + # for lightweight SR (to save parameters) + self.upsample = UpsampleOneStep(upscale, embed_dim, num_out_ch, + (patches_resolution[0], patches_resolution[1])) + elif self.upsampler == 'nearest+conv': + # for real-world SR (less artifacts) + assert self.upscale == 4, 'only support x4 now.' + self.conv_before_upsample = nn.Sequential( + nn.Conv2d(embed_dim, num_feat, 3, 1, 1), nn.LeakyReLU(inplace=True)) + self.conv_up1 = nn.Conv2d(num_feat, num_feat, 3, 1, 1) + self.conv_up2 = nn.Conv2d(num_feat, num_feat, 3, 1, 1) + self.conv_hr = nn.Conv2d(num_feat, num_feat, 3, 1, 1) + self.conv_last = nn.Conv2d(num_feat, num_out_ch, 3, 1, 1) + self.lrelu = nn.LeakyReLU(negative_slope=0.2, inplace=True) + else: + # for image denoising and JPEG compression artifact reduction + self.conv_last = nn.Conv2d(embed_dim, num_out_ch, 3, 1, 1) + + self.apply(self._init_weights) + + def _init_weights(self, m): + if isinstance(m, nn.Linear): + trunc_normal_(m.weight, std=.02) + if isinstance(m, nn.Linear) and m.bias is not None: + nn.init.constant_(m.bias, 0) + elif isinstance(m, nn.LayerNorm): + nn.init.constant_(m.bias, 0) + nn.init.constant_(m.weight, 1.0) + + @torch.jit.ignore + def no_weight_decay(self): + return {'absolute_pos_embed'} + + @torch.jit.ignore + def no_weight_decay_keywords(self): + return {'relative_position_bias_table'} + + def forward_features(self, x): + x_size = (x.shape[2], x.shape[3]) + x = self.patch_embed(x) + if self.ape: + x = x + self.absolute_pos_embed + x = self.pos_drop(x) + + for layer in self.layers: + x = layer(x, x_size) + + x = self.norm(x) # B L C + x = self.patch_unembed(x, x_size) + + return x + + def forward(self, x): + self.mean = self.mean.type_as(x) + x = (x - self.mean) * self.img_range + + if self.upsampler == 'pixelshuffle': + # for classical SR + x = self.conv_first(x) + x = self.conv_after_body(self.forward_features(x)) + x + x = self.conv_before_upsample(x) + x = self.conv_last(self.upsample(x)) + elif self.upsampler == 'pixelshuffledirect': + # for lightweight SR + x = self.conv_first(x) + x = self.conv_after_body(self.forward_features(x)) + x + x = self.upsample(x) + elif self.upsampler == 'nearest+conv': + # for real-world SR + x = self.conv_first(x) + x = self.conv_after_body(self.forward_features(x)) + x + x = self.conv_before_upsample(x) + x = self.lrelu(self.conv_up1(torch.nn.functional.interpolate(x, scale_factor=2, mode='nearest'))) + x = self.lrelu(self.conv_up2(torch.nn.functional.interpolate(x, scale_factor=2, mode='nearest'))) + x = self.conv_last(self.lrelu(self.conv_hr(x))) + else: + # for image denoising and JPEG compression artifact reduction + x_first = self.conv_first(x) + res = self.conv_after_body(self.forward_features(x_first)) + x_first + x = x + self.conv_last(res) + + x = x / self.img_range + self.mean + + return x + + def flops(self): + flops = 0 + H, W = self.patches_resolution + flops += H * W * 3 * self.embed_dim * 9 + flops += self.patch_embed.flops() + for i, layer in enumerate(self.layers): + flops += layer.flops() + flops += H * W * 3 * self.embed_dim * self.embed_dim + flops += self.upsample.flops() + return flops + + +if __name__ == '__main__': + upscale = 4 + window_size = 8 + height = (1024 // upscale // window_size + 1) * window_size + width = (720 // upscale // window_size + 1) * window_size + model = SwinIR( + upscale=2, + img_size=(height, width), + window_size=window_size, + img_range=1., + depths=[6, 6, 6, 6], + embed_dim=60, + num_heads=[6, 6, 6, 6], + mlp_ratio=2, + upsampler='pixelshuffledirect') + print(model) + print(height, width, model.flops() / 1e9) + + x = torch.randn((1, 3, height, width)) + x = model(x) + print(x.shape) diff --git a/basicsr/models/swinir_model.py b/basicsr/models/swinir_model.py new file mode 100644 index 000000000..5ac182f23 --- /dev/null +++ b/basicsr/models/swinir_model.py @@ -0,0 +1,33 @@ +import torch +from torch.nn import functional as F + +from basicsr.utils.registry import MODEL_REGISTRY +from .sr_model import SRModel + + +@MODEL_REGISTRY.register() +class SwinIRModel(SRModel): + + def test(self): + # pad to multiplication of window_size + window_size = self.opt['network_g']['window_size'] + scale = self.opt.get('scale', 1) + mod_pad_h, mod_pad_w = 0, 0 + _, _, h, w = self.lq.size() + if h % window_size != 0: + mod_pad_h = window_size - h % window_size + if w % window_size != 0: + mod_pad_w = window_size - w % window_size + img = F.pad(self.lq, (0, mod_pad_w, 0, mod_pad_h), 'reflect') + if hasattr(self, 'net_g_ema'): + self.net_g_ema.eval() + with torch.no_grad(): + self.output = self.net_g_ema(img) + else: + self.net_g.eval() + with torch.no_grad(): + self.output = self.net_g(img) + self.net_g.train() + + _, _, h, w = self.output.size() + self.output = self.output[:, :, 0:h - mod_pad_h * scale, 0:w - mod_pad_w * scale] diff --git a/docs/HOWTOs.md b/docs/HOWTOs.md index ddda95fde..531b881fb 100644 --- a/docs/HOWTOs.md +++ b/docs/HOWTOs.md @@ -46,3 +46,41 @@ > python inference/inference_dfdnet.py --upscale_factor=2 --test_path datasets/TestWhole 6. The results are in the `results/DFDNet` folder. + +## How to train SwinIR (SR) + +We take the classical SR X4 with DIV2K for example. + +1. Prepare the training dataset: [DIV2K](https://data.vision.ee.ethz.ch/cvl/DIV2K/). More details are in [DatasetPreparation.md](DatasetPreparation.md#image-super-resolution) +1. Prepare the validation dataset: Set5. You can download with [this guidance](DatasetPreparation.md#common-image-sr-datasets) +1. Modify the config file in [`options/train/SwinIR/train_SwinIR_SRx4_scratch.yml`](../options/train/SwinIR/train_SwinIR_SRx4_scratch.yml) accordingly. +1. Train with distributed training. More training commands are in [TrainTest.md](TrainTest.md). + + > python -m torch.distributed.launch --nproc_per_node=8 --master_port=4331 basicsr/train.py -opt options/train/SwinIR/train_SwinIR_SRx4_scratch.yml --launcher pytorch --auto_resume + +Note that: + +1. Different from the original setting in the paper where the X4 model is finetuned from the X2 model, we directly train it from scratch. +1. We also use `EMA (Exponential Moving Average)`. Note that all model trainings in BasicSR supports EMA. +1. In the **250K iteration** of training X4 model, it can achieve comparable performance to the official model. + +| ClassicalSR DIV2KX4 | PSNR (RGB) | PSNR (Y) | SSIM (RGB) | SSIM (Y) | +| :--- | :---: | :---: | :---: | :---: | +| Official | 30.803 | 32.728 | 0.8738|0.9028 | +| Reproduce |30.832 | 32.756 | 0.8739| 0.9025 | + +## How to inference SwinIR (SR) + +1. Download pre-trained models from the [**official SwinIR repo**](https://github.com/JingyunLiang/SwinIR/releases/tag/v0.0) to the `experiments/pretrained_models/SwinIR` folder. +1. Inference. + + > python inference/inference_swinir.py --input datasets/Set5/LRbicx4 --patch_size 48 --model_path experiments/pretrained_models/SwinIR/001_classicalSR_DIV2K_s48w8_SwinIR-M_x4.pth --output results/SwinIR_SRX4_DIV2K/Set5 + +1. The results are in the `results/SwinIR_SRX4_DIV2K/Set5` folder. +1. You may want to calculate the PSNR/SSIM values. + + > python scripts/metrics/calculate_psnr_ssim.py --gt datasets/Set5/GTmod12/ --restored results/SwinIR_SRX4_DIV2K/Set5 --crop_border 4 + + or test with the Y channel with the `--test_y_channel` argument. + + > python scripts/metrics/calculate_psnr_ssim.py --gt datasets/Set5/GTmod12/ --restored results/SwinIR_SRX4_DIV2K/Set5 --crop_border 4 --test_y_channel diff --git a/docs/HOWTOs_CN.md b/docs/HOWTOs_CN.md index df2ab25b8..06472a51f 100644 --- a/docs/HOWTOs_CN.md +++ b/docs/HOWTOs_CN.md @@ -46,3 +46,41 @@ > python inference/inference_dfdnet.py --upscale_factor=2 --test_path datasets/TestWhole 6. 结果在 `results/DFDNet` 文件夹. + +## How to train SwinIR (SR) + +We take the classical SR X4 with DIV2K for example. + +1. Prepare the training dataset: [DIV2K](https://data.vision.ee.ethz.ch/cvl/DIV2K/). More details are in [DatasetPreparation.md](DatasetPreparation.md#image-super-resolution) +1. Prepare the validation dataset: Set5. You can download with [this guidance](DatasetPreparation.md#common-image-sr-datasets) +1. Modify the config file in [`options/train/SwinIR/train_SwinIR_SRx4_scratch.yml`](../options/train/SwinIR/train_SwinIR_SRx4_scratch.yml) accordingly. +1. Train with distributed training. More training commands are in [TrainTest.md](TrainTest.md). + + > python -m torch.distributed.launch --nproc_per_node=8 --master_port=4331 basicsr/train.py -opt options/train/SwinIR/train_SwinIR_SRx4_scratch.yml --launcher pytorch --auto_resume + +Note that: + +1. Different from the original setting in the paper where the X4 model is finetuned from the X2 model, we directly train it from scratch. +1. We also use `EMA (Exponential Moving Average)`. Note that all model trainings in BasicSR supports EMA. +1. In the **250K iteration** of training X4 model, it can achieve comparable performance to the official model. + +| ClassicalSR DIV2KX4 | PSNR (RGB) | PSNR (Y) | SSIM (RGB) | SSIM (Y) | +| :--- | :---: | :---: | :---: | :---: | +| Official | 30.803 | 32.728 | 0.8738|0.9028 | +| Reproduce |30.832 | 32.756 | 0.8739| 0.9025 | + +## How to inference SwinIR (SR) + +1. Download pre-trained models from the [**official SwinIR repo**](https://github.com/JingyunLiang/SwinIR/releases/tag/v0.0) to the `experiments/pretrained_models/SwinIR` folder. +1. Inference. + + > python inference/inference_swinir.py --input datasets/Set5/LRbicx4 --patch_size 48 --model_path experiments/pretrained_models/SwinIR/001_classicalSR_DIV2K_s48w8_SwinIR-M_x4.pth --output results/SwinIR_SRX4_DIV2K/Set5 + +1. The results are in the `results/SwinIR_SRX4_DIV2K/Set5` folder. +1. You may want to calculate the PSNR/SSIM values. + + > python scripts/metrics/calculate_psnr_ssim.py --gt datasets/Set5/GTmod12/ --restored results/SwinIR_SRX4_DIV2K/Set5 --crop_border 4 + + or test with the Y channel with the `--test_y_channel` argument. + + > python scripts/metrics/calculate_psnr_ssim.py --gt datasets/Set5/GTmod12/ --restored results/SwinIR_SRX4_DIV2K/Set5 --crop_border 4 --test_y_channel diff --git a/inference/inference_swinir.py b/inference/inference_swinir.py new file mode 100644 index 000000000..28e9bdeca --- /dev/null +++ b/inference/inference_swinir.py @@ -0,0 +1,199 @@ +# Modified from https://github.com/JingyunLiang/SwinIR +import argparse +import cv2 +import glob +import numpy as np +import os +import torch +from torch.nn import functional as F + +from basicsr.archs.swinir_arch import SwinIR + + +def main(): + parser = argparse.ArgumentParser() + parser.add_argument('--input', type=str, default='datasets/Set5/LRbicx4', help='input test image folder') + parser.add_argument('--output', type=str, default='results/SwinIR/Set5', help='output folder') + parser.add_argument( + '--task', + type=str, + default='classical_sr', + help='classical_sr, lightweight_sr, real_sr, gray_dn, color_dn, jpeg_car') + # dn: denoising; car: compression artifact removal + # TODO: it now only supports sr, need to adapt to dn and jpeg_car + parser.add_argument('--patch_size', type=int, default=64, help='training patch size') + parser.add_argument('--scale', type=int, default=4, help='scale factor: 1, 2, 3, 4, 8') # 1 for dn and jpeg car + parser.add_argument('--noise', type=int, default=15, help='noise level: 15, 25, 50') + parser.add_argument('--jpeg', type=int, default=40, help='scale factor: 10, 20, 30, 40') + parser.add_argument('--large_model', action='store_true', help='Use large model, only used for real image sr') + parser.add_argument( + '--model_path', + type=str, + default='experiments/pretrained_models/SwinIR/001_classicalSR_DF2K_s64w8_SwinIR-M_x4.pth') + args = parser.parse_args() + + os.makedirs(args.output, exist_ok=True) + device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') + # set up model + model = define_model(args) + model.eval() + model = model.to(device) + + if args.task == 'jpeg_car': + window_size = 7 + else: + window_size = 8 + + for idx, path in enumerate(sorted(glob.glob(os.path.join(args.input, '*')))): + # read image + imgname = os.path.splitext(os.path.basename(path))[0] + print('Testing', idx, imgname) + # read image + img = cv2.imread(path, cv2.IMREAD_COLOR).astype(np.float32) / 255. + img = torch.from_numpy(np.transpose(img[:, :, [2, 1, 0]], (2, 0, 1))).float() + img = img.unsqueeze(0).to(device) + + # inference + with torch.no_grad(): + # pad input image to be a multiple of window_size + mod_pad_h, mod_pad_w = 0, 0 + _, _, h, w = img.size() + if h % window_size != 0: + mod_pad_h = window_size - h % window_size + if w % window_size != 0: + mod_pad_w = window_size - w % window_size + img = F.pad(img, (0, mod_pad_w, 0, mod_pad_h), 'reflect') + + output = model(img) + _, _, h, w = output.size() + output = output[:, :, 0:h - mod_pad_h * args.scale, 0:w - mod_pad_w * args.scale] + + # save image + output = output.data.squeeze().float().cpu().clamp_(0, 1).numpy() + if output.ndim == 3: + output = np.transpose(output[[2, 1, 0], :, :], (1, 2, 0)) + output = (output * 255.0).round().astype(np.uint8) + cv2.imwrite(os.path.join(args.output, f'{imgname}_SwinIR.png'), output) + + +def define_model(args): + # 001 classical image sr + if args.task == 'classical_sr': + model = SwinIR( + upscale=args.scale, + in_chans=3, + img_size=args.patch_size, + window_size=8, + img_range=1., + depths=[6, 6, 6, 6, 6, 6], + embed_dim=180, + num_heads=[6, 6, 6, 6, 6, 6], + mlp_ratio=2, + upsampler='pixelshuffle', + resi_connection='1conv') + + # 002 lightweight image sr + # use 'pixelshuffledirect' to save parameters + elif args.task == 'lightweight_sr': + model = SwinIR( + upscale=args.scale, + in_chans=3, + img_size=64, + window_size=8, + img_range=1., + depths=[6, 6, 6, 6], + embed_dim=60, + num_heads=[6, 6, 6, 6], + mlp_ratio=2, + upsampler='pixelshuffledirect', + resi_connection='1conv') + + # 003 real-world image sr + elif args.task == 'real_sr': + if not args.large_model: + # use 'nearest+conv' to avoid block artifacts + model = SwinIR( + upscale=4, + in_chans=3, + img_size=64, + window_size=8, + img_range=1., + depths=[6, 6, 6, 6, 6, 6], + embed_dim=180, + num_heads=[6, 6, 6, 6, 6, 6], + mlp_ratio=2, + upsampler='nearest+conv', + resi_connection='1conv') + else: + # larger model size; use '3conv' to save parameters and memory; use ema for GAN training + model = SwinIR( + upscale=4, + in_chans=3, + img_size=64, + window_size=8, + img_range=1., + depths=[6, 6, 6, 6, 6, 6, 6, 6, 6], + embed_dim=248, + num_heads=[8, 8, 8, 8, 8, 8, 8, 8, 8], + mlp_ratio=2, + upsampler='nearest+conv', + resi_connection='3conv') + + # 004 grayscale image denoising + elif args.task == 'gray_dn': + model = SwinIR( + upscale=1, + in_chans=1, + img_size=128, + window_size=8, + img_range=1., + depths=[6, 6, 6, 6, 6, 6], + embed_dim=180, + num_heads=[6, 6, 6, 6, 6, 6], + mlp_ratio=2, + upsampler='', + resi_connection='1conv') + + # 005 color image denoising + elif args.task == 'color_dn': + model = SwinIR( + upscale=1, + in_chans=3, + img_size=128, + window_size=8, + img_range=1., + depths=[6, 6, 6, 6, 6, 6], + embed_dim=180, + num_heads=[6, 6, 6, 6, 6, 6], + mlp_ratio=2, + upsampler='', + resi_connection='1conv') + + # 006 JPEG compression artifact reduction + # use window_size=7 because JPEG encoding uses 8x8; use img_range=255 because it's slightly better than 1 + elif args.task == 'jpeg_car': + model = SwinIR( + upscale=1, + in_chans=1, + img_size=126, + window_size=7, + img_range=255., + depths=[6, 6, 6, 6, 6, 6], + embed_dim=180, + num_heads=[6, 6, 6, 6, 6, 6], + mlp_ratio=2, + upsampler='', + resi_connection='1conv') + + loadnet = torch.load(args.model_path) + if 'params_ema' in loadnet: + keyname = 'params_ema' + else: + keyname = 'params' + model.load_state_dict(loadnet[keyname], strict=True) + + return model + + +if __name__ == '__main__': + main() diff --git a/options/train/SwinIR/train_SwinIR_SRx2_scratch.yml b/options/train/SwinIR/train_SwinIR_SRx2_scratch.yml new file mode 100644 index 000000000..8402c07ca --- /dev/null +++ b/options/train/SwinIR/train_SwinIR_SRx2_scratch.yml @@ -0,0 +1,106 @@ +# general settings +name: train_SwinIR_SRx2_scratch_P48W8_DIV2K_500k_B4G8 +model_type: SwinIRModel +scale: 2 +num_gpu: auto +manual_seed: 0 + +# dataset and data loader settings +datasets: + train: + name: DIV2K + type: PairedImageDataset + dataroot_gt: datasets/DF2K/DIV2K_train_HR_sub + dataroot_lq: datasets/DF2K/DIV2K_train_LR_bicubic_X2_sub + meta_info_file: basicsr/data/meta_info/meta_info_DIV2K800sub_GT.txt + filename_tmpl: '{}' + io_backend: + type: disk + + gt_size: 96 + use_flip: true + use_rot: true + + # data loader + use_shuffle: true + num_worker_per_gpu: 6 + batch_size_per_gpu: 4 + dataset_enlarge_ratio: 1 + prefetch_mode: ~ + + val: + name: Set5 + type: PairedImageDataset + dataroot_gt: datasets/Set5/GTmod12 + dataroot_lq: datasets/Set5/LRbicx2 + io_backend: + type: disk + +# network structures +network_g: + type: SwinIR + upscale: 2 + in_chans: 3 + img_size: 48 + window_size: 8 + img_range: 1. + depths: [6, 6, 6, 6, 6, 6] + embed_dim: 180 + num_heads: [6, 6, 6, 6, 6, 6] + mlp_ratio: 2 + upsampler: 'pixelshuffle' + resi_connection: '1conv' + +# path +path: + pretrain_network_g: ~ + strict_load_g: true + resume_state: ~ + +# training settings +train: + ema_decay: 0.999 + optim_g: + type: Adam + lr: !!float 2e-4 + weight_decay: 0 + betas: [0.9, 0.99] + + scheduler: + type: MultiStepLR + milestones: [250000, 400000, 450000, 475000] + gamma: 0.5 + + total_iter: 500000 + warmup_iter: -1 # no warm up + + # losses + pixel_opt: + type: L1Loss + loss_weight: 1.0 + reduction: mean + +# validation settings +val: + val_freq: !!float 5e3 + save_img: false + + metrics: + psnr: # metric name, can be arbitrary + type: calculate_psnr + crop_border: 2 + test_y_channel: false + +# logging settings +logger: + print_freq: 100 + save_checkpoint_freq: !!float 5e3 + use_tb_logger: true + wandb: + project: ~ + resume_id: ~ + +# dist training settings +dist_params: + backend: nccl + port: 29500 diff --git a/options/train/SwinIR/train_SwinIR_SRx4_scratch.yml b/options/train/SwinIR/train_SwinIR_SRx4_scratch.yml new file mode 100644 index 000000000..ed1edd006 --- /dev/null +++ b/options/train/SwinIR/train_SwinIR_SRx4_scratch.yml @@ -0,0 +1,106 @@ +# general settings +name: train_SwinIR_SRx4_scratch_P48W8_DIV2K_500k_B4G8 +model_type: SwinIRModel +scale: 4 +num_gpu: auto +manual_seed: 0 + +# dataset and data loader settings +datasets: + train: + name: DIV2K + type: PairedImageDataset + dataroot_gt: datasets/DF2K/DIV2K_train_HR_sub + dataroot_lq: datasets/DF2K/DIV2K_train_LR_bicubic_X4_sub + meta_info_file: basicsr/data/meta_info/meta_info_DIV2K800sub_GT.txt + filename_tmpl: '{}' + io_backend: + type: disk + + gt_size: 192 + use_flip: true + use_rot: true + + # data loader + use_shuffle: true + num_worker_per_gpu: 6 + batch_size_per_gpu: 4 + dataset_enlarge_ratio: 1 + prefetch_mode: ~ + + val: + name: Set5 + type: PairedImageDataset + dataroot_gt: datasets/Set5/GTmod12 + dataroot_lq: datasets/Set5/LRbicx4 + io_backend: + type: disk + +# network structures +network_g: + type: SwinIR + upscale: 4 + in_chans: 3 + img_size: 48 + window_size: 8 + img_range: 1. + depths: [6, 6, 6, 6, 6, 6] + embed_dim: 180 + num_heads: [6, 6, 6, 6, 6, 6] + mlp_ratio: 2 + upsampler: 'pixelshuffle' + resi_connection: '1conv' + +# path +path: + pretrain_network_g: ~ + strict_load_g: true + resume_state: ~ + +# training settings +train: + ema_decay: 0.999 + optim_g: + type: Adam + lr: !!float 2e-4 + weight_decay: 0 + betas: [0.9, 0.99] + + scheduler: + type: MultiStepLR + milestones: [250000, 400000, 450000, 475000] + gamma: 0.5 + + total_iter: 500000 + warmup_iter: -1 # no warm up + + # losses + pixel_opt: + type: L1Loss + loss_weight: 1.0 + reduction: mean + +# validation settings +val: + val_freq: !!float 5e3 + save_img: false + + metrics: + psnr: # metric name, can be arbitrary + type: calculate_psnr + crop_border: 4 + test_y_channel: false + +# logging settings +logger: + print_freq: 100 + save_checkpoint_freq: !!float 5e3 + use_tb_logger: true + wandb: + project: ~ + resume_id: ~ + +# dist training settings +dist_params: + backend: nccl + port: 29500 diff --git a/scripts/metrics/calculate_psnr_ssim.py b/scripts/metrics/calculate_psnr_ssim.py index 39dcecdb8..7f82c0d48 100644 --- a/scripts/metrics/calculate_psnr_ssim.py +++ b/scripts/metrics/calculate_psnr_ssim.py @@ -13,18 +13,22 @@ def main(args): """ psnr_all = [] ssim_all = [] - img_list = sorted(scandir(args.gt, recursive=True, full_path=True)) + img_list_gt = sorted(list(scandir(args.gt, recursive=True, full_path=True))) + img_list_restored = sorted(list(scandir(args.restored, recursive=True, full_path=True))) if args.test_y_channel: print('Testing Y channel.') else: print('Testing RGB channels.') - for i, img_path in enumerate(img_list): + for i, img_path in enumerate(img_list_gt): basename, ext = osp.splitext(osp.basename(img_path)) img_gt = cv2.imread(img_path, cv2.IMREAD_UNCHANGED).astype(np.float32) / 255. - img_restored = cv2.imread(osp.join(args.restored, basename + args.suffix + ext), cv2.IMREAD_UNCHANGED).astype( - np.float32) / 255. + if args.suffix == '': + img_path_restored = img_list_restored[i] + else: + img_path_restored = osp.join(args.restored, basename + args.suffix + ext) + img_restored = cv2.imread(img_path_restored, cv2.IMREAD_UNCHANGED).astype(np.float32) / 255. if args.correct_mean_var: mean_l = [] @@ -63,7 +67,7 @@ def main(args): parser = argparse.ArgumentParser() parser.add_argument('--gt', type=str, default='datasets/val_set14/Set14', help='Path to gt (Ground-Truth)') parser.add_argument('--restored', type=str, default='results/Set14', help='Path to restored images') - parser.add_argument('--crop_border', type=int, default=4, help='Crop border for each side') + parser.add_argument('--crop_border', type=int, default=0, help='Crop border for each side') parser.add_argument('--suffix', type=str, default='', help='Suffix for restored images') parser.add_argument( '--test_y_channel',