Hi,
I want to apply different mask for each image within the same batch using the annotated facial landmarks.
If I understand correctly, I need to first read the fnames within each batch by setting return_fnames=True when creating the data object.
The data pipeline returns me <tf.Tensor 'FIFOQueueV2_DequeueMany:1' shape=(16,) dtype=string>, however I have trouble with viewing/reading this tensor using tf.run(). What would be the correct way to obtain the filenames for each batch?
It returns string tensor and should be easy to view in tf.run() by tf.Print? It's a tf-related question so you might also want to checkout tensorflow communities.
Hi, I want to apply different mask for each image within the same batch using the annotated facial landmarks. If I understand correctly, I need to first read the fnames within each batch by setting
return_fnames=True
when creating the data object.The data pipeline returns me
<tf.Tensor 'FIFOQueueV2_DequeueMany:1' shape=(16,) dtype=string>
, however I have trouble with viewing/reading this tensor using tf.run(). What would be the correct way to obtain the filenames for each batch?Thanks,