UNet backbone. U-Net: Convolutional Networks for Biomedical Image Segmentation. https://arxiv.org/pdf/1505.04597.pdf Args: in_cha

ResNet backbone. Args: depth (int): Depth of resnet, from {18, 34, 50, 101, 152}. in_channels (int): Number of input image channe

Bottleneck block for ResNet. If style is "pytorch", the stride-two layer is the 3x3 conv layer, if it is "caffe", the stride-two layer is the

Fully Convolution Networks for Semantic Segmentation. This head is implemented of `FCNNet <https://arxiv.org/abs/1411.4038>`_. Args:

Base class for BaseDecodeHead. Args: in_channels (int|Sequence[int]): Input channels. channels (int): Channels after modules, bef

Point-wise Spatial Attention Network for Scene Parsing. This head is the implementation of `PSANet <https://hszhao.github.io/papers/eccv18_ps

ResLayer to build ResNet style backbone. Args: block (nn.Module): block used to build ResLayer. inplanes (int): inplanes of block

Upsample convolution block in decoder for UNet. This upsample convolution block consists of one upsample module followed by one convolution b

Accuracy calculation module.

Custom dataset for semantic segmentation. An example of file structure is as followed. .. code-block:: none ├── data │ ├──

Deconvolution upsample module in decoder for UNet (2X upsample). This module uses deconvolution to upsample feature map in the decoder of UNe

InvertedResidual block for MobileNetV2. Args: in_channels (int): The input channels of the InvertedResidual block. out_channels (

Basic convolutional block for UNet. This module consists of several plain convolutional layers. Args: in_channels (int): Number of i

Load an image from file. Required keys are "img_prefix" and "img_info" (a dict that must contain the key "filename"). Added or updated keys a

MobileNetV3 backbone. This backbone is the improved implementation of `Searching for MobileNetV3 <https://ieeexplore.ieee.org/document/900883

CGNet backbone. A Light-weight Context Guided Network for Semantic Segmentation arXiv: https://arxiv.org/abs/1811.08201 Args: in

Context Guided Block for CGNet. This class consists of four components: local feature extractor, surrounding feature extractor, joint feature

Load annotations for semantic segmentation. Args: reduce_zero_label (bool): Whether reduce all label value by 1. Usually used

Interpolation upsample module in decoder for UNet. This module uses interpolation to upsample feature map in the decoder of UNet. It consists

Adaptive Pyramid Context Network for Semantic Segmentation. This head is the implementation of `APCNet <https://openaccess.thecvf.com/content

Basic block for ResNet.

Compose multiple transforms sequentially. Args: transforms (Sequence[dict | callable]): Sequence of transform object or confi

Dynamic Multi-scale Filters for Semantic Segmentation. This head is the implementation of `DMNet <https://openaccess.thecvf.com/content_ICCV_

Inverted Residual Block for MobileNetV3. Args: in_channels (int): The input channels of this Module. out_channels (int): The outp

Rethinking Atrous Convolution for Semantic Image Segmentation. This head is the implementation of `DeepLabV3 <https://arxiv.org/abs/1706.0558

Disentangled Non-Local Neural Networks. This head is the implementation of `DNLNet <https://arxiv.org/abs/2006.06668>`_. Args: r

Distributed evaluation hook. Attributes: dataloader (DataLoader): A PyTorch dataloader. interval (int): Evaluation interval (by e

Evaluation hook. Attributes: dataloader (DataLoader): A PyTorch dataloader. interval (int): Evaluation interval (by epochs). Defa

Linear Embedding

Image to Patch Embedding

Pyramid Scene Parsing Network. This head is the implementation of `PSPNet <https://arxiv.org/abs/1612.01105>`_. Args: pool_scale

Squeeze-and-Excitation Module. Args: channels (int): The input (and output) channels of the SE layer. ratio (int): Squeeze ratio

Unified Perceptual Parsing for Scene Understanding. This head is the implementation of `UPerNet <https://arxiv.org/abs/1807.10221>`_. Ar

A wrapper of concatenated dataset. Same as :obj:`torch.utils.data.dataset.ConcatDataset`, but concat the group flag for image aspect ratio.

Context Encoding for Semantic Segmentation. This head is the implementation of `EncNet <https://arxiv.org/abs/1803.08904>`_. Args:

Lite R-ASPP (LRASPP) head is proposed in Searching for MobileNetV3. This head is the improved implementation of `Searching for MobileNetV3 <h

Online Hard Example Mining Sampler for segmentation. Args: context (nn.Module): The context of sampler, subclass of :obj:`Bas

Pooling Pyramid Module used in PSPNet. Args: pool_scales (tuple[int]): Pooling scales used in Pooling Pyramid Module.

Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation. This head is the implementation of `DeepLabV3+ <https:

Depthwise-Separable Fully Convolutional Network for Semantic Segmentation. This head is implemented according to Fast-SCNN paper. Args:

Fast-SCNN Backbone. Args: in_channels (int): Number of input image channels. Default: 3. downsample_dw_channels (tuple[int]): Num

Global Context Extractor for CGNet. This class is employed to refine the joFint feature of both local feature and surrounding context. A

Downsampling module for CGNet.

A wrapper of repeated dataset. The length of repeated dataset will be `times` larger than the original dataset. This is useful when the data

ResNeSt backbone. Args: groups (int): Number of groups of Bottleneck. Default: 1 base_width (int): Base width of Bottleneck. Defa

ResNeXt backbone. Args: depth (int): Depth of resnet, from {18, 34, 50, 101, 152}. in_channels (int): Number of input image chann

ResNetV1d variant described in [1]_. Compared with default ResNet(ResNetV1b), ResNetV1d replaces the 7x7 conv in the input stem with three 3x

Make a ANN used SelfAttentionBlock. Args: low_in_channels (int): Input channels of lower level feature, which is the key feat

Adaptive Context Module used in APCNet. Args: pool_scale (int): Pooling scale used in Adaptive Context Module to extract regi

Asymmetric Fusion Non-local Block(AFNB) Args: low_in_channels (int): Input channels of lower level feature, which is the key

Asymmetric Non-local Neural Networks for Semantic Segmentation. This head is the implementation of `ANNNet <https://arxiv.org/abs/1908.07678>

Asymmetric Pyramid Non-local Block (APNB) Args: in_channels (int): Input channels of key/query feature, which is the key feat

Atrous Spatial Pyramid Pooling (ASPP) Module. Args: dilations (tuple[int]): Dilation rate of each layer. in_channels (int): Input

Channel Attention Module (CAM)

CCNet: Criss-Cross Attention for Semantic Segmentation. This head is the implementation of `CCNet <https://arxiv.org/abs/1811.11721>`_.

Dual Attention Network for Scene Segmentation. This head is the implementation of `DANet <https://arxiv.org/abs/1809.02983>`_. Args:

Dynamic Convolutional Module used in DMNet. Args: filter_size (int): The filter size of generated convolution kernel used in

Atrous Spatial Pyramid Pooling (ASPP) Module with depthwise separable conv.

Disentangled Non-Local Blocks. Args: temperature (float): Temperature to adjust attention. Default: 0.05

Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).

Expectation Maximization Attention Networks for Semantic Segmentation. This head is the implementation of `EMANet <https://arxiv.org/abs/1907

Expectation Maximization Attention Module used in EMANet. Args: channels (int): Channels of the whole module. num_bases (int): Nu

Encoding Module used in EncNet. Args: in_channels (int): Input channels. num_codes (int): Number of code words. conv_cfg

Encoding Layer: a learnable residual encoder. Input is of shape (batch_size, channels, height, width). Output is of shape (batch_size, num_c

Feature Pyramid Network. This is an implementation of - Feature Pyramid Networks for Object Detection (https://arxiv.org/abs/1612.03144)

Feature fusion module. Args: higher_in_channels (int): Number of input channels of the higher-resolution branch. lowe

GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond. This head is the implementation of `GCNet <https://arxiv.org/abs/1904.

Global feature extractor module. Args: in_channels (int): Number of input channels of the GFE module. Default: 64 blo

High-Resolution Module for HRNet. In this module, every branch has 4 BasicBlocks/Bottlenecks. Fusion/Exchange is in this module.

Learning to downsample module. Args: in_channels (int): Number of input channels. dw_channels (tuple[int]): Number of output chan

A simple pipeline to load image.

Non-local Neural Networks. This head is the implementation of `NLNet <https://arxiv.org/abs/1711.07971>`_. Args: reduction (int)

Object-Contextual Representations for Semantic Segmentation. This head is the implementation of `OCRNet <https://arxiv.org/abs/1909.11065>`_.

Make a OCR used SelfAttentionBlock.

Position Attention Module (PAM) Args: in_channels (int): Input channels of key/query feature. channels (int): Output channels of

Pyramid Pooling Module that only concat the features of each layer. Args: pool_scales (tuple[int]): Pooling scales used in Pooling Pyrami

A mask point head use in PointRend. ``PointHead`` use shared multi-layer perceptron (equivalent to nn.Conv1d) to predict the logit of input p

Radix Softmax module in ``SplitAttentionConv2d``. Args: radix (int): Radix of input. groups (int): Groups of input.

Aggregate the context features according to the initial predicted probability distribution. Employ the soft-weighted method to aggregate the

Split-Attention Conv2d in ResNeSt. Args: in_channels (int): Same as nn.Conv2d. out_channels (int): Same as nn.Conv2d. ker

ADE20K dataset. In segmentation map annotation for ADE20K, 0 stands for background, which is not included in 150 categories. ``reduce_zero_la

Using gamma correction to process the image. Args: gamma (float or int): Gamma value used in gamma correction. Default: 1.0.

Resize images & seg. This class is based on ``Resize``, the only difference is it ensure the long and short sides are divisible by ``siz

Base class for cascade decode head used in :class:`CascadeEncoderDecoder.

Base class of pixel sampler.

Base class for segmentors.

Bottleneck block for ResNeSt. Args: inplane (int): Input planes of this block. planes (int): Middle planes of this block.

Bottleneck block for ResNeXt. If style is "pytorch", the stride-two layer is the 3x3 conv layer, if it is "caffe", the stride-two layer is th

Use CLAHE method to process the image. See `ZUIDERVELD,K. Contrast Limited Adaptive Histogram Equalization[J]. Graphics Gems, 1994:474-485.`

Cascade Encoder Decoder segmentors. CascadeEncoderDecoder almost the same as EncoderDecoder, while decoders of CascadeEncoderDecoder are casc

Center crop the image & seg. Args: crop_size (tuple): Expected size after cropping, (h, w).

Chase_db1 dataset. In segmentation map annotation for Chase_db1, 0 stands for background, which is included in 2 categories. ``reduce_zero_la