tfm.vision.backbones.ResNetUNet

Creates ResNet and ResNet-RS family models.

tfm.vision.backbones.ResNetUNet(
    model_id: int,
    input_specs: tf_keras.layers.InputSpec = layers.InputSpec(shape=[None, None, None, 3]),
    depth_multiplier: float = 1.0,
    stem_type: str = 'v0',
    resnetd_shortcut: bool = False,
    replace_stem_max_pool: bool = False,
    se_ratio: Optional[float] = None,
    init_stochastic_depth_rate: float = 0.0,
    upsample_repeats: Optional[List[int]] = None,
    upsample_filters: Optional[List[int]] = None,
    upsample_kernel_sizes: Optional[List[int]] = None,
    scale_stem: bool = True,
    activation: str = 'relu',
    use_sync_bn: bool = False,
    norm_momentum: float = 0.99,
    norm_epsilon: float = 0.001,
    kernel_initializer: str = 'VarianceScaling',
    kernel_regularizer: Optional[tf_keras.regularizers.Regularizer] = None,
    bias_regularizer: Optional[tf_keras.regularizers.Regularizer] = None,
    bn_trainable: bool = True,
    classification_output: bool = False,
    **kwargs
)

This implements the Deep Residual Network from: Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun. Deep Residual Learning for Image Recognition. (https://arxiv.org/pdf/1512.03385) and Irwan Bello, William Fedus, Xianzhi Du, Ekin D. Cubuk, Aravind Srinivas, Tsung-Yi Lin, Jonathon Shlens, Barret Zoph. Revisiting ResNets: Improved Training and Scaling Strategies. (https://arxiv.org/abs/2103.07579).

model_id An int of the depth of ResNet backbone model.
input_specs A tf_keras.layers.InputSpec of the input tensor.
depth_multiplier A float of the depth multiplier to uniformaly scale up all layers in channel size. This argument is also referred to as width_multiplier in (https://arxiv.org/abs/2103.07579).
stem_type A str of stem type of ResNet. Default to v0. If set to v1, use ResNet-D type stem (https://arxiv.org/abs/1812.01187).
resnetd_shortcut A bool of whether to use ResNet-D shortcut in downsampling blocks.
replace_stem_max_pool A bool of whether to replace the max pool in stem with a stride-2 conv,
se_ratio A float or None. Ratio of the Squeeze-and-Excitation layer.
init_stochastic_depth_rate A float of initial stochastic depth rate.
upsample_repeats A list for upsample repeats of the ConvNext blocks for each level starting from L5, then L4, and so on.
upsample_filters A list for the upsample filter sizes for the ConvNext blocks for each level.
upsample_kernel_sizes A list for upsample kernel sizes for the ConvNext blocks for each level.
scale_stem A bool of whether to scale stem layers.
activation A str name of the activation function.
use_sync_bn If True, use synchronized batch normalization.
norm_momentum A float of normalization momentum for the moving average.
norm_epsilon A small float added to variance to avoid dividing by zero.
kernel_initializer A str for kernel initializer of convolutional layers.
kernel_regularizer A tf_keras.regularizers.Regularizer object for Conv2D. Default to None.
bias_regularizer A tf_keras.regularizers.Regularizer object for Conv2D. Default to None.
bn_trainable A bool that indicates whether batch norm layers should be trainable. Default to True.
classification_output A bool to output the correct level needed for classification (L3), only set to True for pretraining.
**kwargs Additional keyword arguments to be passed.

output_specs A dict of {level: TensorShape} pairs for the model output.

Methods

call

call(
    inputs, training=None, mask=None
)

Calls the model on new inputs and returns the outputs as tensors.

In this case call() just reapplies all ops in the graph to the new inputs (e.g. build a new computational graph from the provided inputs).

Args
inputs Input tensor, or dict/list/tuple of input tensors.
training Boolean or boolean scalar tensor, indicating whether to run the Network in training mode or inference mode.
mask A mask or list of masks. A mask can be either a boolean tensor or None (no mask). For more details, check the guide here.

Returns
A tensor if there is a single output, or a list of tensors if there are more than one outputs.