cross_entropy.py

import tensorflow as tf


def class_balanced_sigmoid_cross_entropy(logits, label, name='cross_entropy_loss'):
    """
    The class-balanced cross entropy loss,
    as in `Holistically-Nested Edge Detection
    <http://arxiv.org/abs/1504.06375>`_.
    This is more numerically stable than class_balanced_cross_entropy

    :param logits: size: the logits.
    :param label: size: the ground truth in {0,1}, of the same shape as logits.
    :returns: a scalar. class-balanced cross entropy loss
    """
    y = tf.cast(label, tf.float32)

    count_neg = tf.reduce_sum(1. - y) # the number of 0 in y
    count_pos = tf.reduce_sum(y) # the number of 1 in y (less than count_neg)
    beta = count_neg / (count_neg + count_pos)

    pos_weight = beta / (1 - beta)
    cost = tf.nn.weighted_cross_entropy_with_logits(logits, y, pos_weight)
    # cost = -tf.reduce_mean(cost * (1 - beta), name=name)
    cost = tf.reduce_mean(cost * (1 - beta), name=name)
    return cost