A residual block that can optionally change the number of channels. :param channels: the number of input channels. :param emb_channels: t

A sequential module that passes timestep embeddings to the children that support it as an extra input.

A general file client to access files in different backend. The client loads a file or text in a specified backend from its path and return i

An attention block that allows spatial positions to attend to each other. Originally ported from here, but adapted to the N-d case. https

Conv and residual block used in BasicVSR. Args: num_in_ch (int): Number of input channels. Default: 3. num_out_ch (int): Number o

This JPEG algorithm result is slightly different from cv2. DiffJPEG supports batch processing. Args: differentiable(bool): If True, u

A downsampling layer with an optional convolution. :param channels: channels in the inputs and outputs. :param use_conv: a bool determini

ResNet block that uses SPADE. It differs from the ResNet block of pix2pixHD in that it takes in the segmentation map as input, learns the ski

Residual block without BN. Args: num_feat (int): Channel number of intermediate features. Default: 64. res_scale (flo

An upsampling layer with an optional convolution. :param channels: channels in the inputs and outputs. :param use_conv: a bool determinin

Conv Layer used in StyleGAN2 Discriminator. Args: in_channels (int): Channel number of the input. out_channels (int): Channel num

The registry that provides name -> object mapping, to support third-party users' custom modules. To create a registry (e.g. a backbone r

Equalized Linear as StyleGAN2. Args: in_channels (int): Size of each sample. out_channels (int): Size of each output sample.

InceptionC block patched for FID computation

Spatial feature transform (SFT) with upsampling block. Args: in_channel (int): Number of input channels. out_channel (int): Numbe

The re-parameterizable block used in the ECBSR architecture. ``Paper: Edge-oriented Convolution Block for Real-time Super Resolution on Mobile De

The model can behave as an encoder (with only self-attention) as well as a decoder, in which case a layer of cross-attention is added between

Conv Layer used in StyleGAN2 Discriminator. Args: in_channels (int): Channel number of the input. out_channels (int): Channel num

The ECB block used in the ECBSR architecture. Paper: Edge-oriented Convolution Block for Real-time Super Resolution on Mobile Devices Ref git

InceptionA block patched for FID computation

Defines the PatchGAN discriminator with the specified arguments.

Residual Dense Block. Used in RRDB block in ESRGAN. Args: num_feat (int): Channel number of intermediate features. num_grow_

Transformer block for image-like data. First, project the input (aka embedding) and reshape to b, t, d. Then apply standard transform

Transformer block for image-like data. First, project the input (aka embedding) and reshape to b, t, d. Then apply standard transform

SpyNet architecture. Args: load_path (str): path for pretrained SpyNet. Default: None.

Style conv. Args: in_channels (int): Channel number of the input. out_channels (int): Channel number of the output. kerne

Style conv. Args: in_channels (int): Channel number of the input. out_channels (int): Channel number of the output. kerne

Upsample, FIR filter, and downsample (smooth version). Args: resample_kernel (list[int]): A list indicating the 1D resample kernel

Modulated deformable conv for deformable alignment. Different from the official DCNv2Pack, which generates offsets and masks from the precedi

Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). From: https://github.com/rwightman/pytorch-image-models

Equalized Linear as StyleGAN2. Args: in_channels (int): Size of each sample. out_channels (int): Size of each output sample.

Data normalization with mean and std. Args: rgb_range (int): Maximum value of RGB. rgb_mean (list[float]): Mean for RGB channels

Modulated Conv2d used in StyleGAN2. There is no bias in ModulatedConv2d. Args: in_channels (int): Channel number of the input.

Modulated Conv2d used in StyleGAN2. There is no bias in ModulatedConv2d. Args: in_channels (int): Channel number of the input.

Alignment module using Pyramid, Cascading and Deformable convolution (PCD). It is used in EDVR. ``Paper: EDVR: Video Restoration with Enhance

r""" Image to Patch Embedding Args: img_size (int): Image size. Default: 224. patch_size (int): Patch token size. Default: 4.

r""" Image to Patch Embedding Args: img_size (int): Image size. Default: 224. patch_size (int): Patch token size. Default: 4.

r""" Image to Patch Unembedding Args: img_size (int): Image size. Default: 224. patch_size (int): Patch token size. Default: 4.

r""" Image to Patch Unembedding Args: img_size (int): Image size. Default: 224. patch_size (int): Patch token size. Default: 4.

A module which performs QKV attention and splits in a different order.

Temporal Spatial Attention (TSA) fusion module. Temporal: Calculate the correlation between center frame and neighboring frames; Spat

To RGB from features. Args: in_channels (int): Channel number of input. num_style_feat (int): Channel number of style features.

To RGB from features. Args: in_channels (int): Channel number of input. num_style_feat (int): Channel number of style features.

Upsample module. Args: scale (int): Scale factor. Supported scales: 2^n and 3. num_feat (int): Channel number of intermediate fea

VGG network for feature extraction. In this implementation, we allow users to choose whether use normalization in the input feature and the t

Uses a pretrained BERT tokenizer by huggingface. Vocab size: 30522 (?)

A basic Swin Transformer layer for one stage. Args: dim (int): Number of input channels. input_resolution (tuple[int]): Input re

A basic Swin Transformer layer for one stage. Args: dim (int): Number of input channels. input_resolution (tuple[int]): Input res

Basic Module for SpyNet.

Basic module of SPyNet. Note that unlike the architecture in spynet_arch.py, the basic module here contains batch normalization.

Construct the embeddings from word and position embeddings.

Merge patches into image

Splitting image into patches

Dequantize CbCr channel

CPU prefetcher. Args: loader: Dataloader.

JPEG Quantization for CbCr channels Args: rounding(function): rounding function to use

CUDA prefetcher. Reference: https://github.com/NVIDIA/apex/issues/304# It may consume more GPU memory. Args: loader: Dataloader

Channel attention. Args: num_feat (int): Channel number of intermediate features. squeeze_factor (int): Channel squeeze factor. D

Channel attention used in RCAN. Args: num_feat (int): Channel number of intermediate features. squeeze_factor (int): Channel sque

Chroma subsampling on CbCr channels

Upsample chroma layers

Full JPEG compression algorithm Args: rounding(function): rounding function to use