vis4d.op.loss.seg_cross_entropy_loss

Segmentation cross entropy loss.

Classes

SegCrossEntropyLoss([reducer])

Segmentation cross entropy loss class.
class SegCrossEntropyLoss(reducer=<function mean_loss>)[source]

Segmentation cross entropy loss class.

Wrapper for nn.CrossEntropyLoss that additionally clips the output to the target size and converts the target mask tensor to long.

__init__(reducer=<function mean_loss>)[source]

Creates an instance of the class.

Parameters:

reducer (LossReducer) – Reducer for the loss function. Defaults to mean_loss.

forward(output, target, ignore_index=255)[source]

Forward pass.

Parameters:

  • output (list[Tensor]) – Model output.

  • target (Tensor) – Assigned segmentation target mask.

  • ignore_index (int) – Ignore class id. Default to 255.

Returns:

Computed loss.

Return type:

LossesType