Source code for mmagic.models.editors.ttsr.ttsr
# Copyright (c) OpenMMLab. All rights reserved.
from typing import Dict, List
import torch
from mmengine.optim import OptimWrapperDict
from mmagic.models.utils import set_requires_grad
from mmagic.registry import MODELS
from mmagic.structures import DataSample
from ..srgan import SRGAN
@MODELS.register_module()
[docs]class TTSR(SRGAN):
"""TTSR model for Reference-based Image Super-Resolution.
Paper: Learning Texture Transformer Network for Image Super-Resolution.
Args:
generator (dict): Config for the generator.
extractor (dict): Config for the extractor.
transformer (dict): Config for the transformer.
pixel_loss (dict): Config for the pixel loss.
discriminator (dict): Config for the discriminator. Default: None.
perceptual_loss (dict): Config for the perceptual loss. Default: None.
transferal_perceptual_loss (dict): Config for the transferal perceptual
loss. Default: None.
gan_loss (dict): Config for the GAN loss. Default: None
train_cfg (dict): Config for train. Default: None.
test_cfg (dict): Config for testing. Default: None.
init_cfg (dict, optional): The weight initialized config for
:class:`BaseModule`. Default: None.
data_preprocessor (dict, optional): The pre-process config of
:class:`BaseDataPreprocessor`. Default: None.
"""
def __init__(self,
generator,
extractor,
transformer,
pixel_loss,
discriminator=None,
perceptual_loss=None,
transferal_perceptual_loss=None,
gan_loss=None,
train_cfg=None,
test_cfg=None,
init_cfg=None,
data_preprocessor=None):
super().__init__(
generator=generator,
discriminator=discriminator,
gan_loss=gan_loss,
pixel_loss=pixel_loss,
perceptual_loss=perceptual_loss,
train_cfg=train_cfg,
test_cfg=test_cfg,
init_cfg=init_cfg,
data_preprocessor=data_preprocessor)
self.transformer = MODELS.build(transformer)
self.extractor = MODELS.build(extractor)
extractor['requires_grad'] = False
self.extractor_copy = MODELS.build(extractor)
set_requires_grad(self.extractor_copy, False)
if transferal_perceptual_loss:
self.transferal_perceptual_loss = MODELS.build(
transferal_perceptual_loss)
else:
self.transferal_perceptual_loss = None
self.pixel_init = train_cfg.get('pixel_init', 0) if train_cfg else 0
[docs] def forward_tensor(self, inputs, data_samples=None, training=False):
"""Forward tensor. Returns result of simple forward.
Args:
inputs (torch.Tensor): batch input tensor collated by
:attr:`data_preprocessor`.
data_samples (List[BaseDataElement], optional):
data samples collated by :attr:`data_preprocessor`.
training (bool): Whether is training. Default: False.
Returns:
(Tensor | Tuple[List[Tensor]]): results of forward inference and
forward train.
"""
img_lq = []
ref_lq = []
ref = []
img_lq = data_samples.img_lq / 255.
ref_lq = data_samples.ref_lq / 255.
ref = data_samples.ref_img / 255.
img_lq, _, _ = self.extractor(img_lq)
ref_lq, _, _ = self.extractor(ref_lq)
refs = self.extractor(ref)
soft_attention, textures = self.transformer(img_lq, ref_lq, refs)
pred = self.generator(inputs, soft_attention, textures)
if training:
return pred, soft_attention, textures
else:
return pred
[docs] def if_run_d(self):
"""Calculates whether need to run the discriminator step."""
return self.step_counter >= self.pixel_init and super().if_run_d()
[docs] def g_step(self, batch_outputs, batch_gt_data: DataSample):
"""G step of GAN: Calculate losses of generator.
Args:
batch_outputs (Tensor): Batch output of generator.
batch_gt_data (Tensor): Batch GT data.
Returns:
dict: Dict of losses.
"""
losses = dict()
pred, soft_attention, textures = batch_outputs
# pix loss
if self.pixel_loss:
losses['loss_pix'] = self.pixel_loss(pred, batch_gt_data)
if self.step_counter >= self.pixel_init:
# perceptual loss
if self.perceptual_loss:
loss_percep, loss_style = self.perceptual_loss(
pred, batch_gt_data)
if loss_percep is not None:
losses['loss_perceptual'] = loss_percep
if loss_style is not None:
losses['loss_style'] = loss_style
# transform loss
if self.transferal_perceptual_loss:
state_dict = self.extractor.module.state_dict() if hasattr(
self.extractor, 'module') else self.extractor.state_dict()
self.extractor_copy.load_state_dict(state_dict)
sr_textures = self.extractor_copy((pred + 1.) / 2.)
losses['loss_transferal'] = self.transferal_perceptual_loss(
sr_textures, soft_attention, textures)
# gan loss for generator
if self.gan_loss and self.discriminator:
fake_g_pred = self.discriminator(pred)
losses['loss_gan'] = self.gan_loss(
fake_g_pred, target_is_real=True, is_disc=False)
return losses
[docs] def g_step_with_optim(self, batch_outputs: torch.Tensor,
batch_gt_data: torch.Tensor,
optim_wrapper: OptimWrapperDict):
"""G step with optim of GAN: Calculate losses of generator and run
optim.
Args:
batch_outputs (Tensor): Batch output of generator.
batch_gt_data (Tensor): Batch GT data.
optim_wrapper (OptimWrapperDict): Optim wrapper dict.
Returns:
dict: Dict of parsed losses.
"""
g_optim_wrapper = optim_wrapper['generator']
e_optim_wrapper = optim_wrapper['extractor']
losses_g = self.g_step(batch_outputs, batch_gt_data)
parsed_losses_g, log_vars_g = self.parse_losses(losses_g)
if g_optim_wrapper.should_update():
g_optim_wrapper.backward(parsed_losses_g)
g_optim_wrapper.step()
g_optim_wrapper.zero_grad()
if e_optim_wrapper.should_update():
e_optim_wrapper.step()
e_optim_wrapper.zero_grad()
return log_vars_g
[docs] def train_step(self, data: List[dict],
optim_wrapper: OptimWrapperDict) -> Dict[str, torch.Tensor]:
"""Train step of GAN-based method.
Args:
data (List[dict]): Data sampled from dataloader.
optim_wrapper (OptimWrapper): OptimWrapper instance
used to update model parameters.
Returns:
Dict[str, torch.Tensor]: A ``dict`` of tensor for logging.
"""
g_optim_wrapper = optim_wrapper['generator']
data = self.data_preprocessor(data, True)
batch_inputs = data['inputs']
data_samples = data['data_samples']
batch_gt_data = self.extract_gt_data(data_samples)
log_vars = dict()
with g_optim_wrapper.optim_context(self):
batch_outputs = self.forward_train(batch_inputs, data_samples)
if self.if_run_g():
set_requires_grad(self.discriminator, False)
log_vars_d = self.g_step_with_optim(
batch_outputs=batch_outputs,
batch_gt_data=batch_gt_data,
optim_wrapper=optim_wrapper)
log_vars.update(log_vars_d)
if self.if_run_d():
set_requires_grad(self.discriminator, True)
pred, _, _ = batch_outputs
for _ in range(self.disc_repeat):
# detach before function call to resolve PyTorch2.0 compile bug
log_vars_d = self.d_step_with_optim(
batch_outputs=pred.detach(),
batch_gt_data=batch_gt_data,
optim_wrapper=optim_wrapper)
log_vars.update(log_vars_d)
if 'loss' in log_vars:
log_vars.pop('loss')
self.step_counter += 1
return log_vars