Computes the Intersection-Over-Union metric for class 0 and/or 1.

Formula:

iou <- true_positives / (true_positives + false_positives + false_negatives)

Intersection-Over-Union is a common evaluation metric for semantic image segmentation.

To compute IoUs, the predictions are accumulated in a confusion matrix, weighted by sample_weight and the metric is then calculated from it.

If sample_weight is NULL, weights default to 1. Use sample_weight of 0 to mask values.

This class can be used to compute IoUs for a binary classification task where the predictions are provided as logits. First a threshold is applied to the predicted values such that those that are below the threshold are converted to class 0 and those that are above the threshold are converted to class 1.

IoUs for classes 0 and 1 are then computed, the mean of IoUs for the classes that are specified by target_class_ids is returned.

Usage

metric_binary_iou(
  ...,
  target_class_ids = list(0L, 1L),
  threshold = 0.5,
  name = NULL,
  dtype = NULL
)

Arguments

...: For forward/backward compatability.
target_class_ids: A list or list of target class ids for which the metric is returned. Options are 0, 1, or c(0, 1). With 0 (or 1), the IoU metric for class 0 (or class 1, respectively) is returned. With c(0, 1), the mean of IoUs for the two classes is returned.
threshold: A threshold that applies to the prediction logits to convert them to either predicted class 0 if the logit is below threshold or predicted class 1 if the logit is above threshold.
name: (Optional) string name of the metric instance.
dtype: (Optional) data type of the metric result.

Value

a Metric instance is returned. The Metric instance can be passed directly to compile(metrics = ), or used as a standalone object. See ?Metric for example usage.

Note

with threshold=0, this metric has the same behavior as IoU.

Examples

Standalone usage:

m <- metric_binary_iou(target_class_ids=c(0L, 1L), threshold = 0.3)
m$update_state(c(0, 1, 0, 1), c(0.1, 0.2, 0.4, 0.7))

## tf.Tensor(
## [[1 1]
##  [1 1]], shape=(2, 2), dtype=int64)

m$result()

## tf.Tensor(0.33333334, shape=(), dtype=float32)

m$reset_state()
m$update_state(c(0, 1, 0, 1), c(0.1, 0.2, 0.4, 0.7),
               sample_weight = 10 * c(0.2, 0.3, 0.4, 0.1))

## tf.Tensor(
## [[2 4]
##  [3 1]], shape=(2, 2), dtype=int64)

m$result()

## tf.Tensor(0.1736111, shape=(), dtype=float32)

Usage with compile() API:

model %>% compile(
    optimizer = 'sgd',
    loss = 'mse',
    metrics = list(metric_binary_iou(
        target_class_ids = 0L,
        threshold = 0.5
    ))
)

Other metrics:
Metric()
custom_metric()
metric_auc()
metric_binary_accuracy()
metric_binary_crossentropy()
metric_binary_focal_crossentropy()
metric_categorical_accuracy()
metric_categorical_crossentropy()
metric_categorical_focal_crossentropy()
metric_categorical_hinge()
metric_concordance_correlation()
metric_cosine_similarity()
metric_f1_score()
metric_false_negatives()
metric_false_positives()
metric_fbeta_score()
metric_hinge()
metric_huber()
metric_iou()
metric_kl_divergence()
metric_log_cosh()
metric_log_cosh_error()
metric_mean()
metric_mean_absolute_error()
metric_mean_absolute_percentage_error()
metric_mean_iou()
metric_mean_squared_error()
metric_mean_squared_logarithmic_error()
metric_mean_wrapper()
metric_one_hot_iou()
metric_one_hot_mean_iou()
metric_pearson_correlation()
metric_poisson()
metric_precision()
metric_precision_at_recall()
metric_r2_score()
metric_recall()
metric_recall_at_precision()
metric_root_mean_squared_error()
metric_sensitivity_at_specificity()
metric_sparse_categorical_accuracy()
metric_sparse_categorical_crossentropy()
metric_sparse_top_k_categorical_accuracy()
metric_specificity_at_sensitivity()
metric_squared_hinge()
metric_sum()
metric_top_k_categorical_accuracy()
metric_true_negatives()
metric_true_positives()