metrics.py

from keras import backend
import tensorflow as tf
import numpy as np


smooth = 1e-15


def dice_coef(y_true, y_pred):
    y_true = tf.keras.layers.Flatten()(y_true)
    y_pred = tf.keras.layers.Flatten()(y_pred)
    intersection = tf.reduce_sum(y_true * y_pred)
    return (2. * intersection + smooth) / (tf.reduce_sum(y_true) + tf.reduce_sum(y_pred) + smooth)


def dice_loss(y_true, y_pred):
    return 1.0 - dice_coef(y_true, y_pred)


def mad(y_true, y_pred):
    return backend.mean(abs(y_pred - y_true))


def iou(y_true, y_pred):
    def f(y_true, y_pred):
        intersection = (y_true * y_pred).sum()
        union = y_true.sum() + y_pred.sum() - intersection
        x = (intersection + smooth) / (union + smooth)
        x = x.astype(np.float32)
        return x
    return tf.numpy_function(f, [y_true, y_pred], tf.float32)