def predict_average(self):
y_lst = []
for i_fold in range(6):
for k in range(18, 19 + 1):
for epoch in range(99, 100 + 1):
y = self.predict(i_fold=i_fold, k=k, epoch=epoch)
y_lst.append(y)
y_avg = np.mean(y_lst, axis=0)
y_avg_bin = self.get_bin(y_avg)
img_clean = self.img_x[..., :3]
y_avg_bin_overlay = semi_transparant(img_clean, y_avg_bin)
concurrent([y_avg[..., 1], y_avg_bin, y_avg_bin_overlay, img_clean])
if 0:
path_save = f'/scratch/lameeus/data/ghent_altar/output/hierarchy/10_lamb/' \
f'paintloss_tiunet_enc{self.fixed_enc}.png'
imsave(path_save, y_avg_bin)
if 1:
path_save = f'/scratch/lameeus/data/ghent_altar/output/hierarchy/10_lamb/' \
f'paintloss_overlay_tiunet_enc{self.fixed_enc}.png'
imsave(path_save, y_avg_bin_overlay)
return 1
def __init__(self):
self.set_data()
if 1:
y_lst = []
for i_fold in range(6):
for k in [12, 13]:
for epoch in [99, 100]:
y = self.predict(i_fold=0, k=k, epoch=epoch)
y_lst.append(y)
y_avg = np.mean(y_lst, axis=0)
y_avg1 = y_avg[..., 1]
plt.imshow(y_avg1)
if 0:
imsave(
f'/home/lameeus/data/ghent_altar/output/hierarchy/13_small/pred_transfer_kfoldenc{self.fixed_enc}_avg.png',
y_avg1)
if 0:
self.predict_average()
if 0:
self.predict_compare()
if 0:
# regular TI-UNet
self.predict_compare_regular()
def folds_annot():
train_data = get_10lamb_old(5)
img_x, _, _, _ = get_training_data(train_data)
img_clean = img_x[..., :3]
lst_get = [get_borders1, get_borders2, get_borders3, get_borders4, get_borders5, get_borders6]
for i_fold in range(6):
img_annot = imread(f'/home/lameeus/data/ghent_altar/input/hierachy/10_lamb/annotations/kfold/annot_{i_fold+1}.png')
y1 = annotations2y(img_annot, thresh=.8)[..., 1]
a = semi_transparant(img_clean, y1.astype(bool))
w0, w1, h0, h1 = lst_get[i_fold]()
clean_annot_crop = a[h0:h1, w0:w1, :]
img_clean_crop = img_clean[h0:h1, w0:w1, :]
if 0: concurrent([img_clean_crop, clean_annot_crop])
folder_save = '/scratch/lameeus/data/ghent_altar/input/hierarchy/10lamb/ifolds'
imsave(os.path.join(folder_save, f'clean_crop_ifold{i_fold}.png'), img_clean_crop)
imsave(os.path.join(folder_save, f'clean_annot_crop_ifold{i_fold}.png'), clean_annot_crop)
pass
def predict_compare_regular(self):
img_y_all = self.k_fold_train_data.get_train_data_all().get_y_train()
lst_get = [
get_borders1, get_borders2, get_borders3, get_borders4,
get_borders5, get_borders6
]
img_clean = self.img_x[..., :3]
for i_fold in range(6):
y_lst = []
# Average prediction
print(f'i_fold = {i_fold}')
for k in [17, 18, 19]:
print(f'k = {k}')
for epoch in [36, 37, 38, 39, 40]:
print(f'epoch = {epoch}')
y = self.predict_regular(i_fold=i_fold, k=k, epoch=epoch)
y_lst.append(y)
y_avg = np.mean(y_lst, axis=0)
y_avg_bin = self.get_bin(y_avg)
# Performance
img_y_te = self.k_fold_train_data.k_split_i(i_fold).get_y_test()
thresh = optimal_test_thresh_equal_distribution(img_y_all, y_avg)
perf = foo_performance(img_y_te, y_avg, thresh)
# CROP
w0, w1, h0, h1 = lst_get[i_fold]()
y_avg_bin_crop = y_avg_bin[h0:h1, w0:w1]
clean_crop = img_clean[h0:h1, w0:w1, :]
y_avg_bin_transparent_crop = semi_transparant(
clean_crop, y_avg_bin_crop)
if 0:
concurrent(
[clean_crop, y_avg_bin_crop, y_avg_bin_transparent_crop])
folder_save = '/scratch/lameeus/data/ghent_altar/output/hierarchy/10_lamb/ifolds_regular_tiunet'
filename = f'_tiunet_ifold{i_fold}_jacc{perf["jaccard"]:.3f}.png'
# Save y_bin
imsave(os.path.join(folder_save, 'binpred' + filename),
y_avg_bin_crop,
b_check_duplicate=False)
# Save overlay
imsave(os.path.join(folder_save, 'overlay' + filename),
y_avg_bin_transparent_crop,
b_check_duplicate=False)
def continues_learning():
folder = '/home/lameeus/data/ghent_altar/input/hierarchy/13_small'
im_clean = imread(os.path.join(folder, 'clean.png'))[..., :3]
im_annot0 = imread(os.path.join(folder, 'annot.tif'))
im_annot1 = imread(os.path.join(folder, 'clean_annot_practical.png'))
y_true = annotations2y(im_annot0)
y_true_extra = annotations2y(im_annot1, thresh=.9)
folder = '/home/lameeus/data/ghent_altar/output/hierarchy/13_small/practical_annotations'
y_pred0 = imread(os.path.join(folder, 'pred_transfer_kfoldenc2_ifold0_avg.png'))
folder = '/home/lameeus/data/ghent_altar/output/hierarchy/13_small'
y_pred1 = imread(os.path.join(folder, 'pred_transfer_kfoldenc2_ifold0_avg_epoch50_J0427.png'))
from performance.testing import optimal_test_thresh_equal_distribution
from scripts.scripts_performance.main_performance import foo_performance
from figures_paper.overlay import semi_transparant
def get_bin(y_pred):
assert len(y_pred.shape) == 2
y_pred01 = np.stack([1-y_pred, y_pred], axis=-1)
thresh = optimal_test_thresh_equal_distribution(y_true, y_pred01)
print(foo_performance(y_true, y_pred01, thresh))
y_pred_bin = y_pred >= thresh
return y_pred_bin
y_pred0_bin = get_bin(y_pred0)
y_pred1_bin = get_bin(y_pred1)
y_pred0_bin_fancy = semi_transparant(im_clean, y_pred0_bin)
y_pred1_bin_fancy = semi_transparant(im_clean, y_pred1_bin)
concurrent([im_clean, y_true[..., 0], y_true_extra[..., 0], y_pred0, y_pred1, y_pred0_bin, y_pred1_bin,
y_pred0_bin_fancy,
y_pred1_bin_fancy])
folder_save = '/home/lameeus/data/ghent_altar/output/hierarchy/13_small/fancy'
imsave(os.path.join(folder_save, 'overalytrain10.png'), y_pred0_bin_fancy)
imsave(os.path.join(folder_save, 'overalytrain10train13.png'), y_pred1_bin_fancy)
return
def combine_stitches():
folder = '/home/lameeus/data/ghent_altar/output/hierarchy/10_lamb/inpainting'
im_base = imread(os.path.join(folder, 'inpainting_stitch_f2_g3_fixed_v0_linear_c256.png'))
im_leftbot = imread(os.path.join(folder, 'inpainting_stitch_f2_g3_fixed_v1_linear_c256_manual.png'))
im_base[3*256:, :(3)*256, :] = im_leftbot[3*256:, :(3)*256, :]
if 0:
plt.imshow(im_base)
imsave(os.path.join(folder, 'inpainting_comb.png'), im_base)
from figures_paper.overlay import semi_transparant
im_clean = imread('/home/lameeus/data/ghent_altar/input/hierarchy/10_lamb/clean.png')
im_paintloss = imread('/home/lameeus/data/ghent_altar/output/hierarchy/10_lamb/detection_updated.png')
im_overlay = semi_transparant(im_clean, im_paintloss)
if 1:
plt.imshow(im_overlay)
folder = '/home/lameeus/data/ghent_altar/output/hierarchy/10_lamb/fancy'
imsave(os.path.join(folder, 'det_overlay.png'), im_overlay)
def lamb():
folder_in = '/net/gaimfs/research/lameeus/data/Nina/'
im_clean = imread(os.path.join(folder_in, 'clean.png'))
im_detection = imread(os.path.join(folder_in, 'paintloss_tiunet_enc0.png'))
extra1 = imread(
'/scratch/lameeus/data/ghent_altar/input/hierarchy/10lamb/mask_pred_comb.png'
)
b1 = np.equal(extra1, 255)
im_update = np.copy(im_clean)
extra2 = imread('/home/lameeus/Desktop/extra_annot.png')
cyan = [0, 255, 255]
im_update[im_detection.astype(bool), :] = cyan
im_update[b1, :] = cyan
y_extra2 = annotations2y(extra2)
y_extra2_0 = y_extra2[..., 0].astype(bool)
y_extra2_1 = y_extra2[..., 1].astype(bool)
im_update[y_extra2_0, :] = im_clean[y_extra2_0, :]
im_update[y_extra2_1, :] = cyan
plt.imshow(im_update)
from data.conversion_tools import detect_colour
paintloss_updated = detect_colour(im_update, 'cyan')
imsave('/net/gaimfs/research/lameeus/data/Nina/detection_updated.png',
paintloss_updated)
return im_update
def foo(n_segm, b=0):
y_pred = n_segm.predict(self.img_x)
thresh_single = optimal_test_thresh_equal_distribution(
self.img_y_te, y_pred)
# data_single_i = {'k': self.k,
# 'i_fold': i_fold,
# 'epoch': epoch}
print(foo_performance(self.img_y_te, y_pred, thresh_single))
img_clean = self.img_x[..., :3]
concurrent([
img_clean, y_pred[..., 1], y_pred[..., 1] >= thresh_single,
semi_transparant(img_clean, y_pred[..., 1] >= thresh_single)
])
if b:
from data.datatools import imsave
folder = '/home/lameeus/data/ghent_altar/output/hierarchy/'
info_epoch = f'_epoch{n_segm.epoch}' if n_segm.epoch > 0 else ''
filename = folder + f'13_small/pred_transfer_kfoldenc{self.fixed_enc}_ifold{self.i_fold}_avg{info_epoch}.png'
imsave(filename, y_pred[..., 1])
resolution_scale = 16
if (modality == 'irr') & (panel_nr == 13):
resolution_scale = 15
im = get_im(panel_nr=panel_nr,
resolution_scale=resolution_scale,
modality=modality,
b_grid=True)
if modality in ('irr', 'ir', 'xray'):
if len(im.shape
) == 3: # In case the grayscale is loaded as 3 colour channels.
im = np.mean(im, axis=2)
if os.name == 'nt': # Laptop windows
if panel_nr == 13:
folder_save = f'C:/Users/Laurens_laptop_w/OneDrive - UGent/data/13_small/webscrape'
else:
folder_save = None
raise NotImplementedError()
else: # Desktop Linux
if 0:
folder_save = f'C:/Users/Laurens_laptop_w/OneDrive - UGent/data/{panel_nr}_small/webscrape'
folder_save = f'/scratch/lameeus/data/ghent_altar/webscrape/{panel_nr}'
assert os.path.exists(folder_save)
path_save = os.path.join(folder_save, f'{modality}_{resolution_scale}.png')
imsave(path_save, im, b_check_duplicate=True)
print("Done")
def predict_compare(self):
img_y_all = self.k_fold_train_data.get_train_data_all().get_y_train()
lst_get = [
get_borders1, get_borders2, get_borders3, get_borders4,
get_borders5, get_borders6
]
img_clean = self.img_x[..., :3]
for i_fold in range(6):
for i_fixed_enc in range(3):
self.fixed_enc = i_fixed_enc
y_lst = []
# Average prediction
for k in [17, 18, 19]:
for epoch in [96, 97, 98, 99, 100]:
y = self.predict(i_fold=i_fold, k=k, epoch=epoch)
y_lst.append(y)
y_avg = np.mean(y_lst, axis=0)
y_avg_bin = self.get_bin(y_avg)
# Performance
img_y_te = self.k_fold_train_data.k_split_i(
i_fold).get_y_test()
thresh = optimal_test_thresh_equal_distribution(
img_y_all, y_avg)
perf = foo_performance(img_y_te, y_avg, thresh)
# CROP
w0, w1, h0, h1 = lst_get[i_fold]()
y_avg_bin_crop = y_avg_bin[h0:h1, w0:w1]
clean_crop = img_clean[h0:h1, w0:w1, :]
y_avg_bin_transparent_crop = semi_transparant(
clean_crop, y_avg_bin_crop)
if 0:
concurrent([
clean_crop, y_avg_bin_crop, y_avg_bin_transparent_crop
])
# Save
if self.fixed_enc == 0: info_enc = 'Train'
elif self.fixed_enc == 1: info_enc = 'Fixed'
elif self.fixed_enc == 2: info_enc = 'FixedTrain'
folder_save = '/scratch/lameeus/data/ghent_altar/output/hierarchy/10_lamb/ifolds'
filename = f'_enc{info_enc}_ifold{i_fold}_jacc{perf["jaccard"]:.3f}.png'
# Save y_bin
imsave(os.path.join(folder_save, 'binpred' + filename),
y_avg_bin_crop,
b_check_duplicate=False)
# Save overlay
imsave(os.path.join(folder_save, 'overlay' + filename),
y_avg_bin_transparent_crop,
b_check_duplicate=False)
"C:/Users/admin/OneDrive - ugentbe/data/images_paper/1319_10nat_V3.png"
), "cyan")
if 0:
plt.imshow(b_mask)
# Read inpaint
im_inpaint = imread(os.path.join(f, "inpainting_comb.jpg"))
im_inpainted_new = inpaint_replacer(im_orig, b_mask, im_inpaint)
plt.imshow(im_inpainted_new)
concurrent([im_orig, im_inpainted_new, im_inpaint, b_mask],
['orig', 'new', 'old', 'mask'])
# Save
imsave(os.path.join(f, "inpaint_v2.png"), im_inpainted_new)
# Crop?
im_treated = imread(os.path.join(f, "restored_SR.jpg"))
# Trying to rescale to similar colour pallet
im_treated_recolour = im_treated.astype(float)
im_treated_recolour = (im_treated_recolour - im_treated.mean(
(0, 1))) * im_inpainted_new.std((0, 1)) / im_treated.std(
(0, 1)) + im_inpainted_new.mean((0, 1))
im_treated_recolour = np.clip(im_treated_recolour, 0, 255)
im_treated_recolour = im_treated_recolour.astype(np.uint8)
imsave(os.path.join(f, "10treated2.png"), im_treated_recolour)
l = []
def combine_annot(b_save=True):
ext = 100
red = [255, 0, 0]
blue = [0, 0, 255]
folder_annots = '/home/lameeus/data/ghent_altar/input/hierachy/10_lamb/annotations/lameeus'
folder_save = '/home/lameeus/data/ghent_altar/input/hierachy/10_lamb/annotations/kfold'
# Clean image
im_clean = imread(
'/home/lameeus/data/ghent_altar/input/hierachy/10_lamb/clean.png')
im_white = 255 * np.ones(shape=im_clean.shape, dtype=np.uint8)
im_annot = np.copy(im_white)
im_annot_clean = np.copy(im_clean)
def apply_redblue(im, h0, h1, w0, w1, b0, b1):
# red to red
# rest to blue
im[h0:h1, w0:w1, :][~b1, :] = blue
im[h0:h1, w0:w1, :][b1, :] = red
for i in range(6):
im = imread(os.path.join(folder_annots,
f'crop{i+1}_annotation_lm.png'))
w0, w1, h0, h1 = lst_get[i]()
im_crop = im[ext:-ext, ext:-ext, :]
y_crop = annotations2y(im_crop, thresh=.9)
b1 = y_crop[..., 1].astype(bool)
if 0:
plt.figure()
plt.imshow(im_crop)
plt.show()
if b_save:
im_annot_i = np.copy(im_white)
im_annot_clean_i = np.copy(im_clean)
apply_redblue(im_annot_i, h0, h1, w0, w1, ~b1, b1)
apply_redblue(im_annot_clean_i, h0, h1, w0, w1, ~b1, b1)
imsave(os.path.join(folder_save, f'annot_{i+1}.png'), im_annot_i)
imsave(os.path.join(folder_save, f'annot_clean_{i+1}.png'),
im_annot_clean_i)
apply_redblue(im_annot, h0, h1, w0, w1, ~b1, b1)
apply_redblue(im_annot_clean, h0, h1, w0, w1, ~b1, b1)
plt.figure()
plt.imshow(im_annot)
plt.show()
if b_save:
imsave(os.path.join(folder_save, 'annot_comb.png'), im_annot)
imsave(os.path.join(folder_save, 'annot_clean_comb.png'),
im_annot_clean)
return
def transfer_learning(
epoch=25, # Could check a few
b_plot=False):
d = 2 # 1, 2
img_x, img_y_val = data_lamb()
k = 10
model_name = 'ti-unet'
w_ext = 10 if d == 1 else 26
# train_data:
y_pred_lst = []
n = ['clean']
# train_data_lst = ['1319_10', '10', '1319', '1319_101319']
train_data_lst = ['10nat', '1319_10nat', '1319_10nat1319', '1319']
data_i_lst = {}
for train_data in train_data_lst:
print(train_data)
epoch_start = 50
epoch_corr = epoch + epoch_start if train_data[:5] == '1319_' else epoch
if train_data == '1319':
epoch_corr = 50
path = f'C:/Users/admin/Data/ghent_altar/net_weight/{train_data}/{model_name}_d{d}_k{k}/w_{epoch_corr}.h5'
try:
model = load_model_quick(path)
except Exception as e:
print(e)
continue
neural_net = NeuralNet(model, w_ext=w_ext, norm_x=True)
y_pred = neural_net.predict(img_x)
# baseline
data_i = _eval_func_single(img_y_val, y_pred, metric='kappa')
print(data_i)
if 0:
"""
Checking which
baseline ~ .22
i = 0: .268, Remove huge improvement ( a lot of "green" background annotated as paint loss)
i = 1: .228 Keep!
i = 2: .179 keep! Drop (keep!!
i = 3: .159 keep! Even more important
i = 4: .252 Remove (huge problem right top)
i = 5: .233 Keep, quit relevant
"""
from datasets.default_trainingsets import get_10lamb_6patches
kFoldTrainData = get_10lamb_6patches(5)
_eval_func_single(
kFoldTrainData.k_split_i(0).get_y_train(),
y_pred,
metric='kappa') # Check what is influence without!
data_i_lst[train_data] = data_i
data_i = _eval_func_single(img_y_val, y_pred, metric='jaccard')
print(data_i)
y_pred_lst.append(y_pred)
n.append(train_data)
# plt.imshow(neural_net.predict(img_x[::2,::2,:])[..., 1])
if b_plot:
concurrent([img_x[..., :3]] + [a[..., 1] for a in y_pred_lst], n)
if 0:
from figures_paper.overlay import semi_transparant
from data.datatools import imread, imsave
t = [data_i_lst[n_i]['thresh'] for n_i in train_data_lst]
p = []
for i, train_data in enumerate(train_data_lst):
b = np.greater_equal(y_pred_lst[i][..., 1], t[i])
k = semi_transparant(img_x[..., :3], b, 0)
p.append(k)
imsave(
os.path.join(
"C:/Users/admin/OneDrive - ugentbe/data/images_paper",
train_data + '.png'), k)
concurrent(p)
return data_i_lst
l = [im_clean]
for i, b in enumerate([b_annot, b_sh, b_pred, b_pred_unet]):
if i == 0:
# Probably not really interesting to change the colour
l.append(
semi_transparant(im_clean,
b.astype(bool),
color1='cyan',
color2='grey',
transparency=0,
transparancy2=.5))
else:
l.append(
semi_transparant(im_clean,
b.astype(bool),
color2='grey',
transparency=0,
transparancy2=.5))
concurrent(l)
folder_out = 'C:/Users/admin/OneDrive - ugentbe/data/images_paper'
imsave(os.path.join(folder_out, f'{data}_clean.png'), im_clean)
imsave(os.path.join(folder_out, f'{data}_annot.png'), l[1])
imsave(os.path.join(folder_out, f'{data}_src.png'), l[2])
imsave(os.path.join(folder_out, f'{data}_tiunet.png'), l[3])
imsave(os.path.join(folder_out, f'{data}_unet.png'), l[4])
