NCSNpp#

class deepinv.models.NCSNpp(img_resolution=64, in_channels=3, out_channels=3, label_dim=0, augment_dim=9, model_channels=128, channel_mult=[1, 2, 2, 2], channel_mult_emb=4, num_blocks=4, attn_resolutions=[16], dropout=0.10, label_dropout=0.0, embedding_type='fourier', channel_mult_noise=2, encoder_type='residual', decoder_type='standard', resample_filter=[1, 3, 3, 1], pretrained=None, pixel_std=0.75, device=None)[source]#

Bases: Denoiser

Re-implementation of the DDPM++ and NCSN++ architectures from the paper: Score-Based Generative Modeling through Stochastic Differential Equations.

Equivalent to the original implementation by Song et al., available at the official implementation.

The model is also pre-conditioned by the method described in the paper Elucidating the Design Space of Diffusion-Based Generative Models.

The architecture consists of a series of convolution layer, down-sampling residual blocks and up-sampling residual blocks with skip-connections of scale \(\sqrt{0.5}\). The model also supports an additional class condition model. Each residual block has a self-attention mechanism with multiple channels per attention head. The noise level can be embedded using either Positional Embedding or Fourier Embedding with optional augmentation linear layer.

Parameters:

  • img_resolution (int) – Image spatial resolution at input/output.

  • in_channels (int) – Number of color channels at input.

  • out_channels (int) – Number of color channels at output.

  • label_dim (int) – Number of class labels, 0 = unconditional.

  • augment_dim (int) – Augmentation label dimensionality, 0 = no augmentation.

  • model_channels (int) – Base multiplier for the number of channels.

  • channel_mult (list) – Per-resolution multipliers for the number of channels.

  • channel_mult_emb (int) – Multiplier for the dimensionality of the embedding vector.

  • num_blocks (int) – Number of residual blocks per resolution.

  • attn_resolutions (list) – List of resolutions with self-attention.

  • dropout (float) – Dropout probability of intermediate activations.

  • label_dropout (float) – Dropout probability of class labels for classifier-free guidance.

  • embedding_type (str) – Timestep embedding type: ‘positional’ for DDPM++, ‘fourier’ for NCSN++.

  • channel_mult_noise (int) – Timestep embedding size: 1 for DDPM++, 2 for NCSN++.

  • encoder_type (str) – Encoder architecture: ‘standard’ for DDPM++, ‘residual’ for NCSN++.

  • decoder_type (str) – Decoder architecture: ‘standard’ for both DDPM++ and NCSN++.

  • resample_filter (list) – Resampling filter: [1,1] for DDPM++, [1,3,3,1] for NCSN++.

  • pretrained (str, None) – use a pretrained network. If pretrained=None, the weights will be initialized at random using Pytorch’s default initialization. If pretrained='download', the weights will be downloaded from an online repository (the default model trained on FFHQ at 64x64 resolution (ffhq64-uncond-ve) with default architecture). Finally, pretrained can also be set as a path to the user’s own pretrained weights. See pretrained-weights for more details.

  • pixel_std (float) – The standard deviation of the normalized pixels (to [0, 1] for example) of the data distribution. Default to 0.75.

  • device (torch.device) – Instruct our module to be either on cpu or on gpu. Default to None, which suggests working on cpu.

forward(x, sigma, class_labels=None, augment_labels=None, *args, **kwargs)[source]#

Run the denoiser on noisy image.

Parameters:
Return torch.Tensor:

denoised image.

forward_unet(x, sigma, class_labels=None, augment_labels=None)[source]#

Run the unet.

Parameters:
Return torch.Tensor:

denoised image.

Examples using NCSNpp:#

Building your diffusion posterior sampling method using SDEs

Building your diffusion posterior sampling method using SDEs