Metrics

This module contains the implementation of the metrics

Dice

class fusionlab.metrics.DiceScore(mode='multiclass', from_logits=True, reduction='none')

__init__(mode='multiclass', from_logits=True, reduction='none')

Computer dice score for binary or multiclass input

Parameters:

• mode – “binary” or “multiclass”

• from_logits – if True, assumes input is raw logits

• reduction – “mean” or “none”, if “none” returns dice score for each channels, else returns mean

forward(y_pred, y_true)

Parameters:

• y_pred – (N, C, *)

• y_true – (N, *)

Return type:

Tensor

Returns:

scalar

fusionlab.metrics.JaccardScore

alias of DiceScore

IoU

class fusionlab.metrics.IoUScore(mode='multiclass', from_logits=True, reduction='none')

__init__(mode='multiclass', from_logits=True, reduction='none')

Implementation of Iou score for segmentation task. It supports “binary”, “multiclass” :param mode: Metric mode {‘binary’, ‘multiclass’} :param from_logits: If True assumes input is raw logits :param reduction: “mean” or “none”, if “none” returns dice score for each channels, else returns mean

forward(y_pred, y_true)

Parameters:

• y_pred – (N, C, *)

• y_true – (N, *)

Returns:

(C, ) if mode is ‘multiclass’ else (1, )