mmagic.models.editors.dcgan

Package Contents

Classes

DCGAN	Implementation of `Unsupervised Representation Learning with Deep
DCGANDiscriminator	Discriminator for DCGAN.
DCGANGenerator	Generator for DCGAN.

class mmagic.models.editors.dcgan.DCGAN(generator: ModelType, discriminator: Optional[ModelType] = None, data_preprocessor: Optional[Union[dict, mmengine.Config]] = None, generator_steps: int = 1, discriminator_steps: int = 1, noise_size: Optional[int] = None, ema_config: Optional[Dict] = None, loss_config: Optional[Dict] = None)

Bases: mmagic.models.base_models.BaseGAN

Implementation of Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks.

Paper link:

<https://arxiv.org/abs/1511.06434>`_ (DCGAN).

Detailed architecture can be found in DCGANGenerator # noqa and DCGANDiscriminator # noqa

disc_loss(disc_pred_fake: torch.Tensor, disc_pred_real: torch.Tensor) → Tuple

Get disc loss. DCGAN use the vanilla gan loss to train the discriminator.

Parameters

• disc_pred_fake (Tensor) – Discriminator’s prediction of the fake images.

• disc_pred_real (Tensor) – Discriminator’s prediction of the real images.

Returns

Loss value and a dict of log variables.

Return type

tuple[Tensor, dict]

gen_loss(disc_pred_fake: torch.Tensor) → Tuple

Get gen loss. DCGAN use the vanilla gan loss to train the generator.

Parameters

disc_pred_fake (Tensor) – Discriminator’s prediction of the fake images.

Returns

Loss value and a dict of log variables.

Return type

tuple[Tensor, dict]

train_discriminator(inputs: dict, data_samples: mmagic.structures.DataSample, optimizer_wrapper: mmengine.optim.OptimWrapper) → Dict[str, torch.Tensor]

Train discriminator.

Parameters

• inputs (dict) – Inputs from dataloader.

• data_samples (DataSample) – Data samples from dataloader.

• optim_wrapper (OptimWrapper) – OptimWrapper instance used to update model parameters.

Returns

A dict of tensor for logging.

Return type

Dict[str, Tensor]

train_generator(inputs: dict, data_samples: mmagic.structures.DataSample, optimizer_wrapper: mmengine.optim.OptimWrapper) → Dict[str, torch.Tensor]

Train generator.

Parameters

• inputs (dict) – Inputs from dataloader.

• data_samples (DataSample) – Data samples from dataloader. Do not used in generator’s training.

• optim_wrapper (OptimWrapper) – OptimWrapper instance used to update model parameters.

Returns

A dict of tensor for logging.

Return type

Dict[str, Tensor]

class mmagic.models.editors.dcgan.DCGANDiscriminator(input_scale, output_scale, out_channels, in_channels=3, base_channels=128, default_norm_cfg=dict(type='BN'), default_act_cfg=dict(type='LeakyReLU'), out_act_cfg=None, init_cfg=None)

Bases: mmengine.model.BaseModule

Discriminator for DCGAN.

Implementation Details for DCGAN architecture:

1. Adopt convolution in the discriminator;

2. Use batchnorm in the discriminator except for the input and final output layer;

3. Use LeakyReLU in the discriminator in addition to the output layer.

Parameters

• input_scale (int) – The scale of the input image.

• output_scale (int) – The final scale of the convolutional feature.

• out_channels (int) – The channel number of the final output layer.

• in_channels (int, optional) – The channel number of the input image. Defaults to 3.

• base_channels (int, optional) – The basic channel number of the generator. The other layers contains channels based on this number. Defaults to 128.

• default_norm_cfg (dict, optional) – Norm config for all of layers except for the final output layer. Defaults to dict(type='BN').

• default_act_cfg (dict, optional) – Activation config for all of layers except for the final output layer. Defaults to dict(type='ReLU').

• out_act_cfg (dict, optional) – Activation config for the final output layer. Defaults to dict(type='Tanh').

• pretrained (str, optional) – Path for the pretrained model. Default to None.

• init_cfg (dict, optional) – Initialization config dict. Default: None.

forward(x)

Forward function.

Parameters

x (torch.Tensor) – Fake or real image tensor.

Returns

Prediction for the reality of the input image.

Return type

torch.Tensor

init_weights()

Init weights for models.

We just use the initialization method proposed in the original paper.

class mmagic.models.editors.dcgan.DCGANGenerator(output_scale, out_channels=3, base_channels=1024, input_scale=4, noise_size=100, default_norm_cfg=dict(type='BN'), default_act_cfg=dict(type='ReLU'), out_act_cfg=dict(type='Tanh'), init_cfg=None)

Bases: mmengine.model.BaseModule

Generator for DCGAN.

Implementation Details for DCGAN architecture:

1. Adopt transposed convolution in the generator;

2. Use batchnorm in the generator except for the final output layer;

3. Use ReLU in the generator in addition to the final output layer.

More details can be found in the original paper: Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks http://arxiv.org/abs/1511.06434

Parameters

• output_scale (int | tuple[int]) – Output scale for the generated image. If only a integer is provided, the output image will be a square shape. The tuple of two integers will set the height and width for the output image, respectively.

• out_channels (int, optional) – The channel number of the output feature. Default to 3.

• base_channels (int, optional) – The basic channel number of the generator. The other layers contains channels based on this number. Default to 1024.

• input_scale (int | tuple[int], optional) – Output scale for the generated image. If only a integer is provided, the input feature ahead of the convolutional generator will be a square shape. The tuple of two integers will set the height and width for the input convolutional feature, respectively. Defaults to 4.

• noise_size (int, optional) – Size of the input noise vector. Defaults to 100.

• default_norm_cfg (dict, optional) – Norm config for all of layers except for the final output layer. Defaults to dict(type='BN').

• default_act_cfg (dict, optional) – Activation config for all of layers except for the final output layer. Defaults to dict(type='ReLU').

• out_act_cfg (dict, optional) – Activation config for the final output layer. Defaults to dict(type='Tanh').

• init_cfg (dict, optional) – Initialization config dict. Default: None.

forward(noise, num_batches=0, return_noise=False)

Forward function.

Parameters

• noise (torch.Tensor | callable | None) – You can directly give a batch of noise through a torch.Tensor or offer a callable function to sample a batch of noise data. Otherwise, the None indicates to use the default noise sampler.

• num_batches (int, optional) – The number of batch size. Defaults to 0.

• return_noise (bool, optional) – If True, noise_batch will be returned in a dict with fake_img. Defaults to False.

Returns

If not return_noise, only the output image

will be returned. Otherwise, a dict contains fake_img and noise_batch will be returned.

Return type

torch.Tensor | dict

init_weights()

Init weights for models.

We just use the initialization method proposed in the original paper.