Focal LOSS说明

Focal LOSS说明
一、理论

二、代码

def sigmoid_focal_loss(y_true, y_pred, gamma=2.0, alpha=0.25):
"""
Compute sigmoid focal loss.
Reference Paper:
"Focal Loss for Dense Object Detection"
https://arxiv.org/abs/1708.02002
# Arguments
y_true: Ground truth targets,
tensor of shape (?, num_boxes, num_classes).
y_pred: Predicted logits,
tensor of shape (?, num_boxes, num_classes).
gamma: exponent of the modulating factor (1 - p_t) ^ gamma.
alpha: optional alpha weighting factor to balance positives vs negatives.
# Returns
sigmoid_focal_loss: Sigmoid focal loss, tensor of shape (?, num_boxes).
"""
sigmoid_loss = K.binary_crossentropy(y_true, y_pred, from_logits=True)
# 预测可能性
pred_prob = tf.sigmoid(y_pred)
# 预测可能是物体的和真实物体的差距，p_t中预测正确的位置权重低，预测错误的权重高
p_t = ((y_true * pred_prob) + ((1 - y_true) * (1 - pred_prob)))
# 对权重做gamma次方，控制权重的大小比例，防止过高或过低
modulating_factor = tf.pow(1.0 - p_t, gamma)
# alpha权重，控制对真实物体的权重较高还是非物体权重较高，alpah越高越重视正样本，
# optional alpha weighting factor to balance positives vs negatives.
alpha_weight_factor = (y_true * alpha + (1 - y_true) * (1 - alpha))

sigmoid_focal_loss = modulating_factor * alpha_weight_factor * sigmoid_loss
return sigmoid_focal_loss