Open nanthakumar94 opened 5 years ago
First remove the suffix ".data-00000-of-00001" from the checkpoint file. So model_fcn8s_final.ckpt.data-00000-of-00001 becomes model_fcn8s_final.ckpt
Then change the sess.restore line to this -
saver.restore(sess, "fcn_8s_checkpoint/model_fcn8s_final.ckpt")
This is my folder order: --fcn_8s_checkpoint --models --nets --preprocessing --recipes --utils
Traceback (most recent call last): File "C:/Users/nanth/PycharmProjects/Background_removal_using_image_segmentation-master/Background_removal_using_image_segmentation-master/segment_image.py", line 59, in
saver.restore(sess, tf.train.latest_checkpoint(checkpoint_dir='model_fcn8s_final.ckpt'))
File "C:\Users\nanth\AppData\Local\Programs\Python\Python36\lib\site-packages\tensorflow\python\training\saver.py", line 1713, in restore
raise ValueError("Can't load save_path when it is None.")
ValueError: Can't load save_path when it is None
please tell me, how to load ckpt file in this script and i need only the segmented image only