Method __init__

monai/networks/nets/resnet.py:369–403

Initialize resnet18 to resnet200 models as a backbone, the backbone can be used as an encoder for segmentation and objection models. Compared with the class `ResNet`, the only different place is the forward function. Args: model_name: name of model to initialize

(self, model_name: str, pretrained: bool = True, spatial_dims: int = 3, in_channels: int = 1)

class ResNetFeatures(ResNet):

    def __init__(self, model_name: str, pretrained: bool = True, spatial_dims: int = 3, in_channels: int = 1) -> None:
        """Initialize resnet18 to resnet200 models as a backbone, the backbone can be used as an encoder for
        segmentation and objection models.

        Compared with the class `ResNet`, the only different place is the forward function.

        Args:
            model_name: name of model to initialize, can be from [resnet10, ..., resnet200].
            pretrained: whether to initialize pretrained MedicalNet weights,
                only available for spatial_dims=3 and in_channels=1.
            spatial_dims: number of spatial dimensions of the input image.
            in_channels: number of input channels for first convolutional layer.
        """
        if model_name not in resnet_params:
            model_name_string = ", ".join(resnet_params.keys())
            raise ValueError(f"invalid model_name {model_name} found, must be one of {model_name_string} ")

        block, layers, shortcut_type, bias_downsample, datasets23 = resnet_params[model_name]

        super().__init__(
            block=block,
            layers=layers,
            block_inplanes=get_inplanes(),
            spatial_dims=spatial_dims,
            n_input_channels=in_channels,
            conv1_t_stride=2,
            shortcut_type=shortcut_type,
            feed_forward=False,
            bias_downsample=bias_downsample,
        )
        if pretrained:
            if spatial_dims == 3 and in_channels == 1:
                _load_state_dict(self, model_name, datasets23=datasets23)
            else:
                raise ValueError("Pretrained resnet models are only available for in_channels=1 and spatial_dims=3.")

    def forward(self, inputs: torch.Tensor):
        """