DeepLabV3+

class mmit.decoders.DeepLabV3Plus(input_channels, input_reductions, decoder_channels=[256, 48], atrous_rates=[12, 24, 36], upsample_layer=<class 'mmit.base.upsamplers.Upsample'>, norm_layer=<class 'torch.nn.modules.batchnorm.BatchNorm2d'>, activation_layer=<class 'torch.nn.modules.activation.ReLU'>, extra_layer=<class 'torch.nn.modules.linear.Identity'>, mismatch_layer=<class 'mmit.base.mismatch.Pad'>, return_features=False)

Implementation of the DeepLabV3+ decoder. Paper: https://arxiv.org/abs/1802.02611 To make it compatible with any encoder, the first skip connection is with the “middle” feature and the second one with the last.

Parameters:

• input_channels (List[int]) – The channels of the input features.

• input_reductions (List[int]) – The reduction factor of the input features.

• decoder_channel – The channel to use on the decoder.

• atrous_rates (List[int]) – The atrous rates to use on the ASPP module.

• norm_layer (Type[Module]) – Normalization layer to use.

• activation_layer (Type[Module]) – Activation function to use.

• extra_layer (Type[Module]) – Addional layer to use.

• return_features (bool) – Whether to return the intermediate results of the decoder.

forward(*features)

Forward pass of the decoder.

Parameters:

*features (Tensor) – Features from the encoder, the first is the input image, last one the deepest.

Return type:

Tensor

property out_classes: int

Number of output classes.