三重损失张量流，其中预先定义了锚正对答案

【问题标题】：Triplet loss tensorflow where anchor-positive pairs are defined upfront三重损失张量流，其中预先定义了锚正对
【发布时间】：2019-04-18 17:56:13
【问题描述】：

我正在尝试在 Tensorflow 中实现三元组损失，其中三元组是通过在线挖掘方式获得的。在我的特殊问题中，我已经有了 anchor(image) - positive(text) 对。我想要实现的是在批处理中有 anchor(image) - positive(text) - negative(text) 和 anchor(text) - positive(image) - negative(image) 的三元组 image-text 对。

如果您需要任何进一步的信息，请告诉我，期待您的回答！

【问题讨论】：

标签： tensorflow deep-learning loss-function triplet

【解决方案1】：

我发现这是我需要的解决方案：

def compute_loss(images: tf.Tensor, texts: tf.Tensor, margin: float) -> tf.Tensor:
    with tf.variable_scope(name_or_scope="loss"):
        scores = tf.matmul(images, texts, transpose_b=True)
        diagonal = tf.diag_part(scores)
        # Compare every diagonal score to scores in its column i.e
        # All contrastive images for each sentence
        cost_s = tf.maximum(0.0, margin - tf.reshape(diagonal, [-1, 1]) + scores)
        # Compare every diagonal score to scores in its row i.e
        # All contrastive sentences for each image
        cost_im = tf.maximum(0.0, margin - diagonal + scores)

        # Clear diagonals
        cost_s = tf.linalg.set_diag(cost_s, tf.zeros(tf.shape(cost_s)[0]))
        cost_im = tf.linalg.set_diag(cost_im, tf.zeros(tf.shape(cost_im)[0]))

        # For each positive pair (i,s) sum over the negative images
        cost_s = tf.reduce_sum(cost_s, axis=1)
        # For each positive pair (i,s) sum over the negative texts
        cost_im = tf.reduce_sum(cost_im, axis=0)

        triplet_loss = tf.reduce_mean(cost_s) + tf.reduce_mean(cost_im)

        return triplet_loss

【讨论】：