tfa.losses.lifted_struct_loss

View source on GitHub

Computes the lifted structured loss.

@tf.function
tfa.losses.lifted_struct_loss(
    labels: tfa.image.color_ops.TensorLike,
    embeddings: tfa.image.color_ops.TensorLike,
    margin: tfa.image.filters.FloatTensorLike = 1.0
) -> tf.Tensor

labels 1-D tf.int32 Tensor with shape [batch_size] of multiclass integer labels.
embeddings 2-D float Tensor of embedding vectors. Embeddings should not be l2 normalized.
margin Float, margin term in the loss definition.

lifted_loss float scalar with dtype of embeddings.