def global_score(image1, image2):
# image1 = "static/images/"+image1
# image2 = "static/images/"+image2
"""this function uses the methods above to calculate a global_score for a couple images.
It returns 1 when we are sure that the two arguments refer to the same image.
Otherwise, it returns an average score."""
try:
if functions.getSize(data_path +
image1) == functions.getSize(data_path + image2):
if SSIM(image1, image2).compare_images()[0] > 0.8:
return 1
elif image1 != image2:
if histogram(image1, image2).correlation() > 0.95:
return 1
else:
h = histogram(image1, image2).correlation()
j = functions.jaccard(image1, image2)
return (h + j) / 2
except Exception as e:
print(444)
print(e)
h = histogram(image1, image2).correlation()
j = functions.jaccard(image1, image2)
return (h + j) / 2
return 0
whole_plot_arr = np.zeros(shape=(len(samples), len(samples)))
for i in range(iterations):
print '\t' + str(
i) + '\t-----------------------------------------------------'
subbed_dat = {}
for dataset in pep_dat:
print '\t' + dataset
subbed_dat[dataset] = fxn.subsample_to_number(
pep_dat, dataset, sample_to)
for x in range(len(samples)):
for y in range(len(samples)):
sampled_jacc[x][y].append(
fxn.jaccard(subbed_dat[samples[x]],
subbed_dat[samples[y]]))
# Generate Venn plots of one examples of subsamples
if i == 0:
fxn.save_venn2(plot_dir + 'sub', subbed_dat, samples[x],
samples[y], samples[x], samples[y])
sub_plot_arr = np.zeros(shape=(len(samples), len(samples)))
for x in range(len(samples)):
for y in range(len(samples)):
sub_plot_arr[x, y] = np.mean(sampled_jacc[x][y])
# Also calculate Jaccards of whole unsampled peptidomes
whole_plot_arr[x, y] = fxn.save_venn2(plot_dir + 'whole', pep_dat,
samples[x], samples[y],
samples[x], samples[y])
def forward(self, y_pr, y_gt):
return 1 - F.jaccard(y_pr, y_gt, eps=self.eps, threshold=None, activation=self.activation)
for i in range(repeats):
print str(
i) + '\t-----------------------------------------------------'
# Subsample each dataset to a fixed number
dat = {}
for dataset in peptides:
print '\t' + dataset
sample = fxn.subsample_to_number(peptides, dataset, sample_to)
dat[dataset] = sample
for x in range(len(pep_samples)):
for y in range(len(pep_samples)):
sampled_jacc[x][y].append(
fxn.jaccard(dat[pep_samples[x]], dat[pep_samples[y]]))
plot_arr = np.zeros(shape=(len(pep_samples), len(pep_samples)))
for x in range(len(pep_samples)):
for y in range(len(pep_samples)):
plot_arr[x, y] = np.mean(sampled_jacc[x][y])
fig = plt.figure(figsize=(10, 8))
ax = fig.add_subplot(111)
p = ax.pcolor(plot_arr, cmap='gnuplot', vmin=0, vmax=1)
ax.set_xticks([x + .5 for x in range(len(pep_samples))])
ax.set_xticklabels(pep_ids, rotation=85)
ax.set_yticks([x + .5 for x in range(len(pep_samples))])
ax.set_yticklabels(pep_ids) # , fontsize=4)
plt.xlim(0, len(pep_samples))
plt.ylim(0, len(pep_samples))
import matplotlib.pyplot as plt
from albumentations.pytorch.transforms import img_to_tensor
arrt = np.random.randn(500, 500)
tensor = img_to_tensor(arrt)
mean=(0.485, 0.456, 0.406)
std=(0.229, 0.224, 0.225)
#new_img = img - mean /std
#old img =
model_path = 'pretrained/unet16_instruments_20/model_1.pt'
model = get_model(model_path, model_type='UNet16', problem_type='instruments')
img_l = load_image('dataset/instrument_dataset_1/left_frames/frame000.png')
input_img_l = torch.unsqueeze(img_to_tensor(img_transform(p=1)(image=img_l)['image']), dim=0)
mask_l = model(input_img_l)
img_r = load_image('dataset/instrument_dataset_1/right_frames/frame000.png')
input_img_r = torch.unsqueeze(img_to_tensor(img_transform(p=1)(image=img_r)['image']), dim=0)
mask_r = model(input_img_r)
mask = jaccard(mask_l, mask_r)
im_seg = mask.data[0].cpu().numpy()[0]
#mask_array = (im_seg * std) + mean
plt.imshow(im_seg > 0)