-
Notifications
You must be signed in to change notification settings - Fork 40
Description
I tried to train another dataset with image size 540x960x3. I also modified INPUT_HEIGHT and INPUT_HEIGHT in config.json file, but it still raised the error:
INFO:tensorflow:Error reported to Coordinator: <class 'tensorflow.python.framework.errors_impl.InvalidArgumentError'>, Input to reshape is a tensor with 1555200 values, but the requested shape has 518400
The error is from this line:
~/Th-Folder/13.SegNet/SegNet-FAT/SegNet.py in train(self, max_steps, batch_size)
226 for step in range(max_steps):
227
--> 228 image_batch, label_batch = self.sess.run([self.images_tr, self.labels_tr])
229
230 feed_dict = {self.inputs_pl: image_batch,
I think the shape 518400 = 360x480x3 is from the original configuration that effects to the input shape request. However, I cannot find the place to change this. Anyone help me?
P/S: I used Tensorflow 1.8.0
Thank you so much!