search

abhaydoke09 / Bilinear-CNN-TensorFlow

This is an implementation of Bilinear CNN for fine grained visual recognition using TensorFlow.
191 stars 72 forks source link

im confused about the img size in train set and validation set #24

Open anqier0468 opened 6 years ago

anqier0468 commented 6 years ago

Actually, i meet this problem. File "/home/..../Bilinear-CNN-TensorFlow/core/bcnn_DD_woft_with_random_crops.py", line 398, in val_loss += sess.run(loss, feed_dict={imgs: batch_val_x, target: batch_val_y}) File "/home/..../python2.7/site-packages/tensorflow/python/client/session.py", line 778, in run run_metadata_ptr) File "/home/....../python2.7/site-packages/tensorflow/python/client/session.py", line 961, in _run % (np_val.shape, subfeed_t.name, str(subfeed_t.get_shape()))) ValueError: Cannot feed value of shape (10, 224, 224, 3) for Tensor u'Placeholder:0', which has shape '(?, 448, 448, 3)' so ,why is the img size of train set and val set, test set different in create_h5_datasets.py?

build_hdf5_image_dataset(new_val, image_shape=(224, 224), mode='file', output_path='new_val_224.h5', categorical_labels=True, normalize=False)' build_hdf5_image_dataset(new_test, image_shape=(224, 224), mode='file', output_path='new_test_224.h5', categorical_labels=True, normalize=False)' build_hdf5_image_dataset(new_train, image_shape=(488, 488), mode='file', output_path='new_train_488.h5', categorical_labels=True, normalize=False)' Hope for your answer. 3Q

JUSTDODoDo commented 6 years ago

I have encountered the same problem. I solved the verification set size to 448x448. I still have a question, can the size of the test set be set to 448x448?