pad=2)
ax.set_aspect('equal')
xstr = np.arange(int(0.004 * 1720) + 1)
x = np.linspace(0, 1720, int(0.004 * 1720) + 1)
plt.yticks(x, xstr)
plt.imshow((data[3]).T, cmap='Greys', vmin=-1, vmax=1, aspect='auto')
plt.title('Residual', fontsize=14, pad=6)
plt.show()
print("Data with size: {}".format(segydata.shape))
chunked_datas = np.expand_dims(
chunker.dimension_preprocess(deepcopy(masked_data)), -1)
chunked_masks = np.expand_dims(chunker.dimension_preprocess(deepcopy(mask)),
-1)
pred_datas = model.predict([chunked_datas, chunked_masks])
chunked_datas = np.squeeze(chunked_datas)
pred_datas = np.squeeze(pred_datas)
reconstructed_data = chunker.dimension_postprocess(pred_datas, segydata)
reconstructed_data = (reconstructed_data * 2 - 1.0) * np.max(abs(segydata_org))
miss_data = deepcopy(segydata_org)
miss_data[mask == 0] = 0
plot_datas([
miss_data, segydata_org, reconstructed_data,
reconstructed_data - segydata_org
])
generator,
steps_per_epoch=2000,
epochs=10,
callbacks=[
TensorBoard(
log_dir='./coco_2017_data/logs/single_image_test',
write_graph=False
),
ModelCheckpoint(
'./coco_2017_data/logs/single_image_test/coco_2017_weights.{epoch:02d}-{loss:.2f}.h5',
monitor='loss',
save_best_only=True,
save_weights_only=True
),
LambdaCallback(
on_epoch_end=lambda epoch, logs: plot_sample_data(
masked_img,
model.predict(
[
np.expand_dims(masked_img,0),
np.expand_dims(mask,0)
]
)[0]
,
img,
middle_title='Prediction'
)
)
],
)
#Generating samples
output_plot = 'predicted_coco_dataset/predicted_rect_valset/'
try:
os.makedirs(output_plot)
except OSError as e:
if e.errno != errno.EEXIST:
raise
n = 0
for (masked, mask), ori, path in tqdm(test_generator):
name = os.path.basename(path)
print(path)
#Run predictions for this batch of new_images
pred_img = model.predict([masked, mask], batch_size=BATCH_SIZE)
pred_time = datetime.datetime.now().strftime('%Y-%m-%d-%H-%M-%S')
#Clear current output and display test images
for i in range(len(ori)):
#for i in range(len(images)):
_, axes = plt.subplots(1, 2, figsize=(10, 5))
axes[0].imshow(masked[i, :, :, :])
axes[1].imshow(pred_img[i, :, :, :] * 1.)
axes[0].set_title('Masked Image')
axes[1].set_title('Predicted Image')
axes[0].xaxis.set_major_formatter(NullFormatter())
axes[0].yaxis.set_major_formatter(NullFormatter())
axes[1].xaxis.set_major_formatter(NullFormatter())
axes[1].yaxis.set_major_formatter(NullFormatter())