HasnainRaz / FC-DenseNet-TensorFlow

Fully Convolutional DenseNet (A.K.A 100 layer tiramisu) for semantic segmentation of images implemented in TensorFlow.
MIT License
123 stars 41 forks source link

seems error of logits reshape when num_classes != 2 #2

Closed LilianYe closed 6 years ago

LilianYe commented 6 years ago

In function xentropy_loss() of model.py logits = tf.reshape(logits, [tf.shape(logits)[0], -1, 2]) should be logits = tf.reshape(logits, [tf.shape(logits)[0], -1, self.num_classes])