def im_show_1(image=None, title='', rows=1, cols=1, index=1):
""" Show image (transfer tensor to numpy first) """
image = normalize_image(image, forward=False)
image = image.permute(1, 2, 0).cpu().numpy()
ax = plt.subplot(rows, cols, index)
ax.set_title(title)
ax.imshow(image.clip(0, 1))
return ax
def vis_feature(vis, model, dataloader, use_cuda=True):
# Visualize feature map of watch image
h = 40
num = 1024
id = 4
for batch_idx, batch in enumerate(dataloader):
if use_cuda:
batch = batch_cuda(batch)
theta, feature_A, feature_B, correlation = model(batch)
if batch_idx == id:
break
watch_feature_A = F.interpolate(feature_A, size=(h, h), mode='bilinear', align_corners=True).transpose(0, 1)[0:num, :, :, :]
watch_feature_B = F.interpolate(feature_B, size=(h, h), mode='bilinear', align_corners=True).transpose(0, 1)[0:num, :, :, :]
opts = dict(jpgquality=100, title='source image')
image_A = normalize_image(batch['source_image'][0], forward=False) * 255.0
vis.image(image_A, opts=opts)
nrow = 32
padding = 3
opts = dict(jpgquality=100, title='feature map A')
# vis.images(watch_feature_A * 255.0, nrow=nrow, padding=padding, opts=opts)
vis.image(torchvision.utils.make_grid(watch_feature_A * 255.0, nrow=nrow, padding=padding), opts=opts)
# vis.image(watch_feature_A[0], opts=opts)
opts = dict(jpgquality=100, title='target image')
image_B = normalize_image(batch['target_image'][0], forward=False) * 255.0
vis.image(image_B, opts=opts)
opts = dict(jpgquality=100, title='feature map B')
vis.image(torchvision.utils.make_grid(watch_feature_B * 255.0, nrow=nrow, padding=padding), opts=opts)
# vis.images(watch_feature_B * 255.0, nrow=nrow, padding=padding, opts=opts)
# vis.image(watch_feature_B[0], opts=opts)
# opts = dict(title='correlation')
# vis.heatmap(correlation[0, 0, :, :], opts=opts)
def draw_image(images, names, flows):
images[:, :, 0:240, :] = normalize_image(images[:, :, 0:240, :],
forward=False)
images *= 255.0
images = images.permute(0, 2, 3, 1).cpu().numpy().astype(np.uint8)
for i in range(images.shape[0]):
pos_name = (80, 255)
if (i + 1) % group_size == 1 or (i + 1) % group_size == 0:
pos_flow = (0, 0)
else:
pos_flow = (70, 275)
cv2.putText(images[i], names[i], pos_name, fnt, 0.5, (0, 0, 0), 1)
cv2.putText(images[i], flows[i], pos_flow, fnt, 0.5, (0, 0, 0), 1)
images = torch.Tensor(images.astype(np.float32))
images = images.permute(0, 3, 1, 2)
return images
def vis_pf(vis,
dataloader,
theta,
theta_weak,
theta_inver,
theta_weak_inver,
results,
results_weak,
dataset_name,
use_cuda=True):
# Visualize watch images
affTnf = GeometricTnf(geometric_model='affine', use_cuda=use_cuda)
tpsTnf = GeometricTnf(geometric_model='tps', use_cuda=use_cuda)
pt = PointTnf(use_cuda=use_cuda)
watch_images = torch.ones(len(dataloader) * 6, 3, 280, 240)
watch_keypoints = -torch.ones(len(dataloader) * 6, 2, 20)
if use_cuda:
watch_images = watch_images.cuda()
watch_keypoints = watch_keypoints.cuda()
num_points = np.ones(len(dataloader) * 6).astype(np.int8)
correct_index = list()
image_names = list()
metrics = list()
# Colors for keypoints
cmap = plt.get_cmap('tab20')
colors = list()
for c in range(20):
r = cmap(c)[0] * 255
g = cmap(c)[1] * 255
b = cmap(c)[2] * 255
colors.append((b, g, r))
fnt = cv2.FONT_HERSHEY_COMPLEX
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32))
for batch_idx, batch in enumerate(dataloader):
if use_cuda:
batch = batch_cuda(batch)
# Theta and theta_inver
theta_aff = theta['aff'][batch_idx].unsqueeze(0)
theta_aff_tps = theta['aff_tps'][batch_idx].unsqueeze(0)
theta_weak_aff = theta_weak['aff'][batch_idx].unsqueeze(0)
theta_weak_aff_tps = theta_weak['aff_tps'][batch_idx].unsqueeze(0)
theta_aff_inver = theta_inver['aff'][batch_idx].unsqueeze(0)
theta_aff_tps_inver = theta_inver['aff_tps'][batch_idx].unsqueeze(0)
theta_weak_aff_inver = theta_weak_inver['aff'][batch_idx].unsqueeze(0)
theta_weak_aff_tps_inver = theta_weak_inver['aff_tps'][
batch_idx].unsqueeze(0)
# Warped image
warped_aff = affTnf(batch['source_image'], theta_aff)
warped_aff_tps = tpsTnf(warped_aff, theta_aff_tps)
warped_weak_aff = affTnf(batch['source_image'], theta_weak_aff)
warped_weak_aff_tps = tpsTnf(warped_weak_aff, theta_weak_aff_tps)
watch_images[batch_idx * 6, :, 0:240, :] = batch['source_image']
watch_images[batch_idx * 6 + 1, :, 0:240, :] = warped_aff
watch_images[batch_idx * 6 + 2, :, 0:240, :] = warped_aff_tps
watch_images[batch_idx * 6 + 3, :, 0:240, :] = batch['target_image']
watch_images[batch_idx * 6 + 4, :, 0:240, :] = warped_weak_aff
watch_images[batch_idx * 6 + 5, :, 0:240, :] = warped_weak_aff_tps
# Warped keypoints
source_im_size = batch['source_im_info'][:, 0:3]
target_im_size = batch['target_im_info'][:, 0:3]
source_points = batch['source_points']
target_points = batch['target_points']
source_points_norm = PointsToUnitCoords(P=source_points,
im_size=source_im_size)
target_points_norm = PointsToUnitCoords(P=target_points,
im_size=target_im_size)
warped_points_aff_norm = pt.affPointTnf(theta=theta_aff_inver,
points=source_points_norm)
warped_points_aff = PointsToPixelCoords(P=warped_points_aff_norm,
im_size=target_im_size)
pck_aff, index_aff, N_pts = pck(target_points, warped_points_aff,
dataset_name)
warped_points_aff = relocate(warped_points_aff, target_im_size)
warped_points_aff_tps_norm = pt.tpsPointTnf(theta=theta_aff_tps_inver,
points=source_points_norm)
warped_points_aff_tps_norm = pt.affPointTnf(
theta=theta_aff_inver, points=warped_points_aff_tps_norm)
warped_points_aff_tps = PointsToPixelCoords(
P=warped_points_aff_tps_norm, im_size=target_im_size)
pck_aff_tps, index_aff_tps, _ = pck(target_points,
warped_points_aff_tps,
dataset_name)
warped_points_aff_tps = relocate(warped_points_aff_tps, target_im_size)
warped_points_weak_aff_norm = pt.affPointTnf(
theta=theta_weak_aff_inver, points=source_points_norm)
warped_points_weak_aff = PointsToPixelCoords(
P=warped_points_weak_aff_norm, im_size=target_im_size)
pck_weak_aff, index_weak_aff, _ = pck(target_points,
warped_points_weak_aff,
dataset_name)
warped_points_weak_aff = relocate(warped_points_weak_aff,
target_im_size)
warped_points_weak_aff_tps_norm = pt.tpsPointTnf(
theta=theta_weak_aff_tps_inver, points=source_points_norm)
warped_points_weak_aff_tps_norm = pt.affPointTnf(
theta=theta_weak_aff_inver, points=warped_points_weak_aff_tps_norm)
warped_points_weak_aff_tps = PointsToPixelCoords(
P=warped_points_weak_aff_tps_norm, im_size=target_im_size)
pck_weak_aff_tps, index_weak_aff_tps, _ = pck(
target_points, warped_points_weak_aff_tps, dataset_name)
warped_points_weak_aff_tps = relocate(warped_points_weak_aff_tps,
target_im_size)
watch_keypoints[batch_idx * 6, :, :N_pts] = relocate(
batch['source_points'], source_im_size)[:, :, :N_pts]
watch_keypoints[batch_idx * 6 +
1, :, :N_pts] = warped_points_aff[:, :, :N_pts]
watch_keypoints[batch_idx * 6 +
2, :, :N_pts] = warped_points_aff_tps[:, :, :N_pts]
watch_keypoints[batch_idx * 6 + 3, :, :N_pts] = relocate(
batch['target_points'], target_im_size)[:, :, :N_pts]
watch_keypoints[batch_idx * 6 +
4, :, :N_pts] = warped_points_weak_aff[:, :, :N_pts]
watch_keypoints[
batch_idx * 6 +
5, :, :N_pts] = warped_points_weak_aff_tps[:, :, :N_pts]
num_points[batch_idx * 6:batch_idx * 6 + 6] = N_pts
correct_index.append(np.arange(N_pts))
correct_index.append(index_aff)
correct_index.append(index_aff_tps)
correct_index.append(np.arange(N_pts))
correct_index.append(index_weak_aff)
correct_index.append(index_weak_aff_tps)
image_names.append('Source')
image_names.append('Aff')
image_names.append('Aff_tps')
image_names.append('Target')
image_names.append('Rocco_aff')
image_names.append('Rocco_aff_tps')
metrics.append('')
metrics.append('PCK: {:.2%}'.format(pck_aff))
metrics.append('PCK: {:.2%}'.format(pck_aff_tps))
metrics.append('')
metrics.append('PCK: {:.2%}'.format(pck_weak_aff))
metrics.append('PCK: {:.2%}'.format(pck_weak_aff_tps))
opts = dict(jpgquality=100, title=dataset_name)
# Un-normalize for caffe resnet and vgg
# watch_images = watch_images.permute(0, 2, 3, 1) + pixel_means
# watch_images = watch_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
# watch_images = normalize_image(watch_images, forward=False) * 255.0
watch_images[:, :, 0:240, :] = normalize_image(watch_images[:, :,
0:240, :],
forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3,
1).cpu().numpy().astype(np.uint8)
watch_keypoints = watch_keypoints.cpu().numpy()
for i in range(watch_images.shape[0]):
pos_name = (80, 255)
if (i + 1) % 6 == 1 or (i + 1) % 6 == 4:
pos_pck = (0, 0)
else:
pos_pck = (70, 275)
cv2.putText(watch_images[i], image_names[i], pos_name, fnt, 0.5,
(0, 0, 0), 1)
cv2.putText(watch_images[i], metrics[i], pos_pck, fnt, 0.5, (0, 0, 0),
1)
if (i + 1) % 6 == 4:
for j in range(num_points[i]):
cv2.drawMarker(
watch_images[i],
(watch_keypoints[i, 0, j], watch_keypoints[i, 1, j]),
colors[j], cv2.MARKER_DIAMOND, 12, 2, cv2.LINE_AA)
else:
for j in correct_index[i]:
cv2.drawMarker(
watch_images[i],
(watch_keypoints[i, 0, j], watch_keypoints[i, 1, j]),
colors[j], cv2.MARKER_CROSS, 12, 2, cv2.LINE_AA)
cv2.drawMarker(watch_images[i],
(watch_keypoints[i + 3 - (i % 6), 0, j],
watch_keypoints[i + 3 - (i % 6), 1, j]),
colors[j], cv2.MARKER_DIAMOND, 12, 2,
cv2.LINE_AA)
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images, nrow=3, padding=3),
opts=opts)
def vis_caltech(vis,
dataloader,
theta,
theta_weak,
results,
results_weak,
title,
use_cuda=True):
# Visualize watch images
affTnf = GeometricTnf(geometric_model='affine', use_cuda=use_cuda)
tpsTnf = GeometricTnf(geometric_model='tps', use_cuda=use_cuda)
watch_images = torch.ones(len(dataloader) * 6, 3, 280, 240)
if use_cuda:
watch_images = watch_images.cuda()
image_names = list()
lt_acc = list()
iou = list()
fnt = cv2.FONT_HERSHEY_COMPLEX
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32))
for batch_idx, batch in enumerate(dataloader):
if use_cuda:
batch = batch_cuda(batch)
theta_aff = theta['aff'][batch_idx].unsqueeze(0)
theta_aff_tps = theta['aff_tps'][batch_idx].unsqueeze(0)
theta_weak_aff = theta_weak['aff'][batch_idx].unsqueeze(0)
theta_weak_aff_tps = theta_weak['aff_tps'][batch_idx].unsqueeze(0)
# Warped image
warped_aff = affTnf(batch['source_image'], theta_aff)
warped_aff_tps = tpsTnf(warped_aff, theta_aff_tps)
warped_weak_aff = affTnf(batch['source_image'], theta_weak_aff)
warped_weak_aff_tps = tpsTnf(warped_weak_aff, theta_weak_aff_tps)
watch_images[batch_idx * 6, :, 0:240, :] = batch['source_image']
watch_images[batch_idx * 6 + 1, :, 0:240, :] = warped_aff
watch_images[batch_idx * 6 + 2, :, 0:240, :] = warped_aff_tps
watch_images[batch_idx * 6 + 3, :, 0:240, :] = batch['target_image']
watch_images[batch_idx * 6 + 4, :, 0:240, :] = warped_weak_aff
watch_images[batch_idx * 6 + 5, :, 0:240, :] = warped_weak_aff_tps
image_names.append('Source')
image_names.append('Aff')
image_names.append('Aff_tps')
image_names.append('Target')
image_names.append('Rocco_aff')
image_names.append('Rocco_aff_tps')
lt_acc.append('')
lt_acc.append('LT-ACC: {:.2f}'.format(
float(results['aff']['label_transfer_accuracy'][batch_idx])))
lt_acc.append('LT-ACC: {:.2f}'.format(
float(results['aff_tps']['label_transfer_accuracy'][batch_idx])))
lt_acc.append('')
lt_acc.append('LT-ACC: {:.2f}'.format(
float(results_weak['aff']['label_transfer_accuracy'][batch_idx])))
lt_acc.append('LT-ACC: {:.2f}'.format(
float(results_weak['aff_tps']['label_transfer_accuracy']
[batch_idx])))
iou.append('')
iou.append('IoU: {:.2f}'.format(
float(results['aff']['intersection_over_union'][batch_idx])))
iou.append('IoU: {:.2f}'.format(
float(results['aff_tps']['intersection_over_union'][batch_idx])))
iou.append('')
iou.append('IoU: {:.2f}'.format(
float(results_weak['aff']['intersection_over_union'][batch_idx])))
iou.append('IoU: {:.2f}'.format(
float(results_weak['aff_tps']['intersection_over_union']
[batch_idx])))
opts = dict(jpgquality=100, title=title)
# Un-normalize for caffe resnet and vgg
# watch_images = watch_images.permute(0, 2, 3, 1) + pixel_means
# watch_images = watch_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
# watch_images = normalize_image(watch_images, forward=False) * 255.0
watch_images[:, :, 0:240, :] = normalize_image(watch_images[:, :,
0:240, :],
forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3,
1).cpu().numpy().astype(np.uint8)
for i in range(watch_images.shape[0]):
pos_name = (80, 255)
if (i + 1) % 6 == 1 or (i + 1) % 6 == 4:
pos_lt_ac = (0, 0)
pos_iou = (0, 0)
else:
pos_lt_ac = (10, 275)
pos_iou = (140, 275)
cv2.putText(watch_images[i], image_names[i], pos_name, fnt, 0.5,
(0, 0, 0), 1)
cv2.putText(watch_images[i], lt_acc[i], pos_lt_ac, fnt, 0.5, (0, 0, 0),
1)
cv2.putText(watch_images[i], iou[i], pos_iou, fnt, 0.5, (0, 0, 0), 1)
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images, nrow=3, padding=5),
opts=opts)
def vis_fn(vis, train_loss, val_pck, train_lr, epoch, num_epochs, dataloader, theta, thetai, results,
geometric_model='tps', use_cuda=True):
# Visualize watch images
if geometric_model == 'tps':
geoTnf = GeometricTnf(geometric_model='tps', use_cuda=use_cuda)
elif geometric_model == 'affine':
geoTnf = GeometricTnf(geometric_model='affine', use_cuda=use_cuda)
pt = PointTPS(use_cuda=use_cuda)
group_size = 3
watch_images = torch.ones(len(dataloader) * group_size, 3, 280, 240)
watch_keypoints = -torch.ones(len(dataloader) * group_size, 2, 20)
if use_cuda:
watch_images = watch_images.cuda()
watch_keypoints = watch_keypoints.cuda()
num_points = np.ones(len(dataloader) * group_size).astype(np.int8)
correct_index = list()
image_names = list()
metrics = list()
# Colors for keypoints
cmap = plt.get_cmap('tab20')
colors = list()
for c in range(20):
r = cmap(c)[0] * 255
g = cmap(c)[1] * 255
b = cmap(c)[2] * 255
colors.append((b, g, r))
fnt = cv2.FONT_HERSHEY_COMPLEX
theta, thetai = swap(theta, thetai)
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32))
for batch_idx, batch in enumerate(dataloader):
if use_cuda:
batch = batch_cuda(batch)
batch['source_image'], batch['target_image'] = swap(batch['source_image'], batch['target_image'])
batch['source_im_info'], batch['target_im_info'] = swap(batch['source_im_info'], batch['target_im_info'])
batch['source_points'], batch['target_points'] = swap(batch['source_points'], batch['target_points'])
# Theta and thetai
if geometric_model == 'tps':
theta_batch = theta['tps'][batch_idx].unsqueeze(0)
theta_batch_inver = thetai['tps'][batch_idx].unsqueeze(0)
elif geometric_model == 'affine':
theta_batch = theta['aff'][batch_idx].unsqueeze(0)
theta_batch_inver = thetai['aff'][batch_idx].unsqueeze(0)
# Warped image
warped_image = geoTnf(batch['source_image'], theta_batch)
watch_images[batch_idx * group_size, :, 0:240, :] = batch['source_image']
watch_images[batch_idx * group_size + 1, :, 0:240, :] = warped_image
watch_images[batch_idx * group_size + 2, :, 0:240, :] = batch['target_image']
# Warped keypoints
source_im_size = batch['source_im_info'][:, 0:3]
target_im_size = batch['target_im_info'][:, 0:3]
source_points = batch['source_points']
target_points = batch['target_points']
source_points_norm = PointsToUnitCoords(P=source_points, im_size=source_im_size)
target_points_norm = PointsToUnitCoords(P=target_points, im_size=target_im_size)
if geometric_model == 'tps':
warped_points_norm = pt.tpsPointTnf(theta=theta_batch_inver, points=source_points_norm)
elif geometric_model == 'affine':
warped_points_norm = pt.affPointTnf(theta=theta_batch_inver, points=source_points_norm)
warped_points = PointsToPixelCoords(P=warped_points_norm, im_size=target_im_size)
_, index_correct, N_pts = pck(target_points, warped_points)
watch_keypoints[batch_idx * group_size, :, :N_pts] = relocate(batch['source_points'], source_im_size)[:, :, :N_pts]
watch_keypoints[batch_idx * group_size + 1, :, :N_pts] = relocate(warped_points, target_im_size)[:, :, :N_pts]
watch_keypoints[batch_idx * group_size + 2, :, :N_pts] = relocate(batch['target_points'], target_im_size)[:, :, :N_pts]
num_points[batch_idx * group_size:batch_idx * group_size + group_size] = N_pts
correct_index.append(np.arange(N_pts))
correct_index.append(index_correct)
correct_index.append(np.arange(N_pts))
image_names.append('Source')
if geometric_model == 'tps':
image_names.append('TPS')
elif geometric_model == 'affine':
image_names.append('Affine')
image_names.append('Target')
metrics.append('')
if geometric_model == 'tps':
metrics.append('PCK: {:.2%}'.format(float(results['tps']['pck'][batch_idx])))
elif geometric_model == 'affine':
metrics.append('PCK: {:.2%}'.format(float(results['aff']['pck'][batch_idx])))
metrics.append('')
opts = dict(jpgquality=100, title='Epoch ' + str(epoch) + ' source warped target')
# Un-normalize for caffe resnet and vgg
# watch_images = watch_images.permute(0, 2, 3, 1) + pixel_means
# watch_images = watch_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
# watch_images = normalize_image(watch_images, forward=False) * 255.0
watch_images[:, :, 0:240, :] = normalize_image(watch_images[:, :, 0:240, :], forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3, 1).cpu().numpy().astype(np.uint8)
watch_keypoints = watch_keypoints.cpu().numpy()
for i in range(watch_images.shape[0]):
pos_name = (80, 255)
if (i + 1) % group_size == 1 or (i + 1) % group_size == 0:
pos_pck = (0, 0)
else:
pos_pck = (70, 275)
cv2.putText(watch_images[i], image_names[i], pos_name, fnt, 0.5, (0, 0, 0), 1)
cv2.putText(watch_images[i], metrics[i], pos_pck, fnt, 0.5, (0, 0, 0), 1)
if (i + 1) % group_size == 0:
for j in range(num_points[i]):
cv2.drawMarker(watch_images[i], (watch_keypoints[i, 0, j], watch_keypoints[i, 1, j]), colors[j],
cv2.MARKER_CROSS, 12, 2, cv2.LINE_AA)
else:
for j in correct_index[i]:
cv2.drawMarker(watch_images[i], (watch_keypoints[i, 0, j], watch_keypoints[i, 1, j]), colors[j],
cv2.MARKER_DIAMOND, 12, 2, cv2.LINE_AA)
cv2.drawMarker(watch_images[i],
(watch_keypoints[i + (group_size - 1) - (i % group_size), 0, j], watch_keypoints[i + (group_size - 1) - (i % group_size), 1, j]),
colors[j], cv2.MARKER_CROSS, 12, 2, cv2.LINE_AA)
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images, nrow=6, padding=3), opts=opts)
if epoch == num_epochs:
if geometric_model == 'affine':
sub_str = 'Affine'
elif geometric_model == 'tps':
sub_str = 'TPS'
epochs = np.arange(1, num_epochs+1)
# Visualize train loss
opts_loss = dict(xlabel='Epoch',
ylabel='Loss',
title='GM ResNet101 ' + sub_str + ' Training Loss',
legend=['Loss'],
width=2000)
vis.line(train_loss, epochs, opts=opts_loss)
# Visualize val pck
opts_pck = dict(xlabel='Epoch',
ylabel='Val PCK',
title='GM ResNet101 ' + sub_str + ' Val PCK',
legend=['PCK'],
width=2000)
vis.line(val_pck, epochs, opts=opts_pck)
# Visualize train lr
opts_lr = dict(xlabel='Epoch',
ylabel='Learning Rate',
title='GM ResNet101 ' + sub_str + ' Training Learning Rate',
legend=['LR'],
width=2000)
vis.line(train_lr, epochs, opts=opts_lr)
def train_fn(epoch,
model,
loss_fn,
loss_cycle_fn,
loss_jitter_fn,
lambda_c,
lambda_j,
optimizer,
dataloader,
triple_generation,
geometric_model='tps',
use_cuda=True,
log_interval=100,
vis=None):
"""
Train the model with synthetically training triple:
{source image, target image, refer image (warped source image), theta_GT} from PF-PASCAL.
1. Train the transformation parameters theta_st from source image to target image;
2. Train the transformation parameters theta_tr from target image to refer image;
3. Combine theta_st and theta_st to obtain theta from source image to refer image, and compute loss between
theta and theta_GT.
"""
geoTnf = GeometricTnf(geometric_model=geometric_model, use_cuda=use_cuda)
epoch_loss = 0
if (epoch % 5 == 0 or epoch == 1) and vis is not None:
stride_images = len(dataloader) / 3
group_size = 9
watch_images = torch.ones(group_size * 4, 3, 340, 340).cuda()
watch_theta = torch.zeros(8, 36).cuda()
fnt = cv2.FONT_HERSHEY_COMPLEX
stride_loss = len(dataloader) / 105
iter_loss = np.zeros(106)
begin = time.time()
for batch_idx, batch in enumerate(dataloader):
''' Move input batch to gpu '''
# batch['source_image'].shape & batch['target_image'].shape: (batch_size, 3, 240, 240)
# batch['theta'].shape-tps: (batch_size, 18)-random or (batch_size, 18, 1, 1)-(pre-set from csv)
if use_cuda:
batch = batch_cuda(batch)
''' Get the training triple {source image, target image, refer image (warped source image), theta_GT}'''
batch_triple = triple_generation(batch)
''' Train the model '''
optimizer.zero_grad()
loss = 0
# Predict tps parameters between images
# theta.shape: (batch_size, 18) for tps, theta.shape: (batch_size, 6) for affine
theta_st, theta_ts, theta_tr, theta_rt = model(
batch_triple) # from source image to target image
loss_match = loss_fn(theta_st=theta_st,
theta_tr=theta_tr,
theta_GT=batch_triple['theta_GT'])
loss_cycle_st = loss_cycle_fn(theta_AB=theta_st, theta_BA=theta_ts)
loss_cycle_ts = loss_cycle_fn(theta_AB=theta_ts, theta_BA=theta_st)
loss_cycle_tr = loss_cycle_fn(theta_AB=theta_tr, theta_BA=theta_rt)
loss_cycle_rt = loss_cycle_fn(theta_AB=theta_rt, theta_BA=theta_tr)
loss_jitter = loss_jitter_fn(theta_st=theta_st, theta_tr=theta_tr)
loss = loss_match + lambda_c * (
loss_cycle_st + loss_cycle_ts + loss_cycle_tr +
loss_cycle_rt) / 4 + lambda_j * loss_jitter
# loss = loss_match + lambda_c * (loss_cycle_st + loss_cycle_ts + loss_cycle_tr + loss_cycle_rt) / 4
loss.backward()
optimizer.step()
epoch_loss += loss.item()
if (batch_idx + 1) % log_interval == 0:
end = time.time()
print(
'Train epoch: {} [{}/{} ({:.0%})]\t\tCurrent batch loss: {:.6f}\t\tTime cost ({} batches): {:.4f} s'
.format(epoch, batch_idx + 1,
len(dataloader), (batch_idx + 1) / len(dataloader),
loss.item(), batch_idx + 1, end - begin))
if (epoch % 5 == 0 or epoch == 1) and vis is not None:
if (batch_idx + 1) % stride_images == 0 or batch_idx == 0:
watch_images, watch_theta = add_watch(
watch_images, watch_theta, batch_triple, geoTnf, theta_st,
theta_tr,
int((batch_idx + 1) / stride_images) * group_size,
int((batch_idx + 1) / stride_images))
# if batch_idx <= 19:
# watch_images, image_names = add_watch(watch_images, watch_theta, batch_st, batch_tr, geoTnf, theta_st, theta_tr, batch_idx * group_size, batch_idx)
# if batch_idx == 19:
# opts = dict(jpgquality=100, title='Epoch ' + str(epoch) + ' source warped_sr target warped_tr refer warped_sr')
# watch_images[:, :, 50:290, 50:290] = normalize_image(watch_images[:, :, 50:290, 50:290], forward=False)
# watch_images *= 255.0
# watch_images = watch_images.permute(0, 2, 3, 1).cpu().numpy().astype(np.uint8)
# for i in range(watch_images.shape[0]):
# if i % group_size == 0:
# cp_norm = watch_theta[int(i / group_size), :18].view(1, 2, -1)
# watch_images[i] = draw_grid(watch_images[i], cp_norm)
#
# cp_norm = watch_theta[int(i / group_size), 18:].view(1, 2, -1)
# watch_images[i + 1] = draw_grid(watch_images[i + 1], cp_norm)
#
# cp_norm = watch_theta[int(i / group_size) + 1, :18].view(1, 2, -1)
# watch_images[i + 3] = draw_grid(watch_images[i + 3], cp_norm)
#
# cp_norm = watch_theta[int(i / group_size) + 1, 18:].view(1, 2, -1)
# watch_images[i + 4] = draw_grid(watch_images[i + 4], cp_norm)
#
# watch_images = torch.Tensor(watch_images.astype(np.float32))
# watch_images = watch_images.permute(0, 3, 1, 2)
# vis.image(torchvision.utils.make_grid(watch_images, nrow=3, padding=3), opts=opts)
if (batch_idx + 1) % stride_loss == 0 or batch_idx == 0:
iter_loss[int((batch_idx + 1) /
stride_loss)] = epoch_loss / (batch_idx + 1)
# watch_images = normalize_image(batch_tr['target_image'], forward=False) * 255.0
# vis.images(watch_images, nrow=8, padding=3)
end = time.time()
# Visualize watch images & train loss
if (epoch % 5 == 0 or epoch == 1) and vis is not None:
opts = dict(jpgquality=100,
title='Epoch ' + str(epoch) +
' source warped_sr target warped_tr refer warped_sr')
watch_images[:, :, 50:290,
50:290] = normalize_image(watch_images[:, :, 50:290,
50:290],
forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3,
1).cpu().numpy().astype(np.uint8)
for i in range(watch_images.shape[0]):
if i % group_size == 0:
cp_norm = watch_theta[int(i / group_size), :18].view(1, 2, -1)
watch_images[i] = draw_grid(watch_images[i], cp_norm)
cp_norm = watch_theta[int(i / group_size), 18:].view(1, 2, -1)
watch_images[i + 1] = draw_grid(watch_images[i + 1], cp_norm)
cp_norm = watch_theta[int(i / group_size) + 1, :18].view(
1, 2, -1)
watch_images[i + 3] = draw_grid(watch_images[i + 3], cp_norm)
cp_norm = watch_theta[int(i / group_size) + 1,
18:].view(1, 2, -1)
watch_images[i + 4] = draw_grid(watch_images[i + 4], cp_norm)
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images, nrow=3, padding=3),
opts=opts)
opts_loss = dict(xlabel='Iterations (' + str(stride_loss) + ')',
ylabel='Loss',
title='GM ResNet101 ' + geometric_model +
' Training Loss in Epoch ' + str(epoch),
legend=['Loss'],
width=2000)
vis.line(iter_loss, np.arange(106), opts=opts_loss)
epoch_loss /= len(dataloader)
print('Train set -- Average loss: {:.6f}\t\tTime cost: {:.4f}'.format(
epoch_loss, end - begin))
return epoch_loss, end - begin
def train_fn(epoch,
model,
loss_fn,
optimizer,
dataloader,
use_cuda=True,
log_interval=100,
vis=None):
"""
Train cosegmentation model:
{source image, target image} from PF-PASCAL.
1. Train the co-object masks on source image and target image;
2. Compute loss.
"""
epoch_loss = 0
# if (epoch % 5 == 0 or epoch == 1) and vis is not None:
if vis is not None:
stride_images = len(dataloader) / 4
group_size = 6
watch_images = torch.ones(group_size * 5, 3, 260, 240).cuda()
image_names = list()
fnt = cv2.FONT_HERSHEY_COMPLEX
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32)).cuda()
stride_loss = len(dataloader) / 92
iter_loss = np.zeros(93)
begin = time.time()
for batch_idx, batch in enumerate(dataloader):
''' Move input batch to gpu '''
if use_cuda:
batch = batch_cuda(batch)
''' Train the model '''
optimizer.zero_grad()
mask_A, mask_B = model(batch)
loss = loss_fn(mask_A, mask_B, batch['source_image'],
batch['target_image'])
loss.backward()
optimizer.step()
epoch_loss += loss.item()
if (batch_idx + 1) % log_interval == 0:
end = time.time()
print(
'Train epoch: {} [{}/{} ({:.0%})]\t\tCurrent batch loss: {:.6f}\t\tTime cost ({} batches): {:.4f} s'
.format(epoch, batch_idx + 1,
len(dataloader), (batch_idx + 1) / len(dataloader),
loss.item(), batch_idx + 1, end - begin))
# if (epoch % 5 == 0 or epoch == 1) and vis is not None:
if vis is not None:
if (batch_idx + 1) % stride_images == 0 or batch_idx == 0:
watch_images, image_names = add_watch(
watch_images, image_names, batch, mask_A, mask_B,
int((batch_idx + 1) / stride_images) * group_size)
# if batch_idx <= 4:
# watch_images, image_names = add_watch(watch_images, image_names, batch, mask_A, mask_B, batch_idx * group_size)
# if batch_idx == 4:
# opts = dict(jpgquality=100, title='Epoch ' + str(epoch) + ' image mask')
# watch_images[:, :, 0:240, :] = normalize_image(watch_images[:, :, 0:240, :], forward=False)
# watch_images *= 255.0
# watch_images = watch_images.permute(0, 2, 3, 1).cpu().numpy().astype(np.uint8)
# for i in range(len(image_names)):
# cv2.putText(watch_images[i], image_names[i], (80, 255), fnt, 0.5, (0, 0, 0), 1)
# watch_images = torch.Tensor(watch_images.astype(np.float32))
# watch_images = watch_images.permute(0, 3, 1, 2)
# vis.image(torchvision.utils.make_grid(watch_images, nrow=group_size, padding=3), opts=opts)
if (batch_idx + 1) % stride_loss == 0 or batch_idx == 0:
iter_loss[int((batch_idx + 1) /
stride_loss)] = epoch_loss / (batch_idx + 1)
end = time.time()
# Visualize watch images & train loss
# if (epoch % 5 == 0 or epoch == 1) and vis is not None:
if vis is not None:
opts = dict(jpgquality=100,
title='Epoch ' + str(epoch) + ' image mask')
watch_images[:, :, 0:240, :] = normalize_image(watch_images[:, :,
0:240, :],
forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3,
1).cpu().numpy().astype(np.uint8)
for i in range(watch_images.shape[0]):
cv2.putText(watch_images[i], image_names[i], (80, 255), fnt, 0.5,
(0, 0, 0), 1)
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images,
nrow=group_size,
padding=3),
opts=opts)
# Un-normalize for caffe resnet and vgg
# watch_images = watch_images.permute(0, 2, 3, 1) + pixel_means
# watch_images = watch_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
opts_loss = dict(xlabel='Iterations (' + str(stride_loss) + ')',
ylabel='Loss',
title='CoSegmentation Training Loss in Epoch ' +
str(epoch),
legend=['Loss'],
width=2000)
vis.line(iter_loss, np.arange(93), opts=opts_loss)
epoch_loss /= len(dataloader)
print('Train set -- Average loss: {:.6f}\t\tTime cost: {:.4f}'.format(
epoch_loss, end - begin))
return epoch_loss, end - begin
def vis_fn(vis,
train_loss,
val_iou,
train_lr,
epoch,
num_epochs,
dataloader,
results,
masks_A,
masks_B,
use_cuda=True):
# Visualize watch images
group_size = 6
watch_images = torch.ones(len(dataloader) * group_size, 3, 280, 240)
if use_cuda:
watch_images = watch_images.cuda()
image_names = list()
metrics = list()
# Colors for keypoints
cmap = plt.get_cmap('tab20')
colors = list()
for c in range(20):
r = cmap(c)[0] * 255
g = cmap(c)[1] * 255
b = cmap(c)[2] * 255
colors.append((b, g, r))
fnt = cv2.FONT_HERSHEY_COMPLEX
for batch_idx, batch in enumerate(dataloader):
if use_cuda:
batch = batch_cuda(batch)
# Theta and theta_inver
watch_images[batch_idx * group_size, :,
0:240, :] = batch['source_image']
watch_images[batch_idx * group_size + 1, :,
0:240, :] = torch.mul(batch['source_image'],
batch['source_mask'])
# watch_images[batch_idx * group_size + 2, :, 0:240, :] = torch.mul(batch['source_image'], masks_A[batch_idx])
mask_A = masks_A[batch_idx].gt(0.5).float()
watch_images[batch_idx * group_size + 2, :,
0:240, :] = torch.mul(batch['source_image'], mask_A)
watch_images[batch_idx * group_size + 3, :,
0:240, :] = batch['target_image']
watch_images[batch_idx * group_size + 4, :,
0:240, :] = torch.mul(batch['target_image'],
batch['target_mask'])
# watch_images[batch_idx * group_size + 5, :, 0:240, :] = torch.mul(batch['target_image'], masks_B[batch_idx])
mask_B = masks_B[batch_idx].gt(0.5).float()
watch_images[batch_idx * group_size + 5, :,
0:240, :] = torch.mul(batch['target_image'], mask_B)
image_names.append('Source')
image_names.append('Mask_GT')
image_names.append('Mask')
image_names.append('Target')
image_names.append('Mask_GT')
image_names.append('Mask')
metrics.append('')
metrics.append('')
metrics.append('IoU: {:.2%}'.format(float(results[batch_idx, 0])))
metrics.append('')
metrics.append('')
metrics.append('IoU: {:.2%}'.format(float(results[batch_idx, 1])))
opts = dict(jpgquality=100,
title='Epoch ' + str(epoch) + ' image mask_gt mask')
# Un-normalize for caffe resnet and vgg
# watch_images = watch_images.permute(0, 2, 3, 1) + pixel_means
# watch_images = watch_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
# watch_images = normalize_image(watch_images, forward=False) * 255.0
watch_images[:, :, 0:240, :] = normalize_image(watch_images[:, :,
0:240, :],
forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3,
1).cpu().numpy().astype(np.uint8)
for i in range(watch_images.shape[0]):
pos_name = (80, 255)
if (i + 1) % group_size == 3 or (i + 1) % group_size == 0:
pos_iou = (70, 275)
else:
pos_iou = (0, 0)
cv2.putText(watch_images[i], image_names[i], pos_name, fnt, 0.5,
(0, 0, 0), 1)
cv2.putText(watch_images[i], metrics[i], pos_iou, fnt, 0.5, (0, 0, 0),
1)
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images, nrow=6, padding=3),
opts=opts)
if epoch == num_epochs:
epochs = np.arange(1, num_epochs + 1)
# Visualize train loss
opts_loss = dict(xlabel='Epoch',
ylabel='Loss',
title='CoSegmentation Training Loss',
legend=['Loss'],
width=2000)
vis.line(train_loss, epochs, opts=opts_loss)
# Visualize val pck
opts_pck = dict(xlabel='Epoch',
ylabel='Val PCK',
title='CoSegmentation Val IoU',
legend=['PCK'],
width=2000)
vis.line(val_iou, epochs, opts=opts_pck)
# Visualize train lr
opts_lr = dict(xlabel='Epoch',
ylabel='Learning Rate',
title='CoSegmentation Training Learning Rate',
legend=['LR'],
width=2000)
vis.line(train_lr, epochs, opts=opts_lr)
def vis_fn(vis,
train_loss,
val_pck,
train_lr,
epoch,
num_epochs,
dataloader,
theta,
thetai,
results,
geometric_model='tps',
use_cuda=True):
geoTnf = GeometricTnf(geometric_model='tps', use_cuda=use_cuda)
group_size = 3
watch_images = torch.ones(len(dataloader) * group_size, 3, 340, 340)
if use_cuda:
watch_images = watch_images.cuda()
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32))
for batch_idx, batch in enumerate(dataloader):
if use_cuda:
batch = batch_cuda(batch)
# Theta and thetai
theta_batch = theta['tps'][batch_idx].unsqueeze(0)
# Warped image
warped_image = geoTnf(batch['source_image'], theta_batch)
watch_images[batch_idx * group_size, :, 50:290,
50:290] = batch['source_image']
watch_images[batch_idx * group_size + 1, :, 50:290,
50:290] = warped_image
watch_images[batch_idx * group_size + 2, :, 50:290,
50:290] = batch['target_image']
opts = dict(jpgquality=100,
title='Epoch ' + str(epoch) + ' source warped target')
# Un-normalize for caffe resnet and vgg
# watch_images = watch_images.permute(0, 2, 3, 1) + pixel_means
# watch_images = watch_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
# watch_images = normalize_image(watch_images, forward=False) * 255.0
watch_images[:, :, 50:290,
50:290] = normalize_image(watch_images[:, :, 50:290, 50:290],
forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3,
1).cpu().numpy().astype(np.uint8)
for i in range(watch_images.shape[0]):
if i % group_size == 0:
cp_norm = theta['tps'][int(i / group_size)][:18].view(1, 2, -1)
watch_images[i] = draw_grid(watch_images[i], cp_norm)
cp_norm = theta['tps'][int(i / group_size)][18:].view(1, 2, -1)
watch_images[i + 1] = draw_grid(watch_images[i + 1], cp_norm)
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images, nrow=3, padding=3),
opts=opts)
if epoch == num_epochs:
if geometric_model == 'affine':
sub_str = 'Affine'
elif geometric_model == 'tps':
sub_str = 'TPS'
epochs = np.arange(1, num_epochs + 1)
# Visualize train loss
opts_loss = dict(xlabel='Epoch',
ylabel='Loss',
title='GM ResNet101 ' + sub_str + ' Training Loss',
legend=['Loss'],
width=2000)
vis.line(train_loss, epochs, opts=opts_loss)
# Visualize val pck
opts_pck = dict(xlabel='Epoch',
ylabel='Val PCK',
title='GM ResNet101 ' + sub_str + ' Val PCK',
legend=['PCK'],
width=2000)
vis.line(val_pck, epochs, opts=opts_pck)
# Visualize train lr
opts_lr = dict(xlabel='Epoch',
ylabel='Learning Rate',
title='GM ResNet101 ' + sub_str +
' Training Learning Rate',
legend=['LR'],
width=2000)
vis.line(train_lr, epochs, opts=opts_lr)
def train_fn_detect(epoch,
model,
faster_rcnn,
aff_theta,
loss_fn,
optimizer,
dataloader,
triple_generation,
use_cuda=True,
log_interval=100,
vis=None,
show=False):
"""
Train the model with synthetically training triple:
{source image, target image, refer image (warped source image), theta_GT} from PF-PASCAL.
1. Train the transformation parameters theta_st from source image to target image;
2. Train the transformation parameters theta_tr from target image to refer image;
3. Combine theta_st and theta_st to obtain theta from source image to refer image, and compute loss between
theta and theta_GT.
"""
tpsTnf = GeometricTnf(geometric_model='tps', use_cuda=use_cuda)
epoch_loss = 0
if (epoch % 5 == 0 or epoch == 1) and vis is not None:
stride_images = len(dataloader) / 3
watch_images = torch.Tensor(24, 3, 240, 240)
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32)).cuda()
stride_loss = len(dataloader) / 35
iter_loss = np.zeros(36)
begin = time.time()
for batch_idx, batch in enumerate(dataloader):
''' Move input batch to gpu '''
# batch['source_image'].shape & batch['target_image'].shape: (batch_size, 3, 240, 240)
# batch['theta'].shape-tps: (batch_size, 18)-random or (batch_size, 18, 1, 1)-(pre-set from csv)
if use_cuda:
batch = batch_cuda(batch)
''' Get the training triple {source image, target image, refer image (warped source image), theta_GT}'''
batch_triple = triple_generation(batch)
''' Train the model '''
optimizer.zero_grad()
# Predict tps parameters between images
box_info_s = faster_rcnn(
im_data=batch_triple['source_im'],
im_info=batch_triple['source_im_info'][:, 3:],
gt_boxes=batch_triple['source_gt_boxes'],
num_boxes=batch_triple['source_num_boxes'])[0:3]
box_info_t = faster_rcnn(
im_data=batch_triple['target_im'],
im_info=batch_triple['target_im_info'][:, 3:],
gt_boxes=batch_triple['target_gt_boxes'],
num_boxes=batch_triple['target_num_boxes'])[0:3]
box_info_r = faster_rcnn(
im_data=batch_triple['refer_im'],
im_info=batch_triple['refer_im_info'][:, 3:],
gt_boxes=batch_triple['refer_gt_boxes'],
num_boxes=batch_triple['refer_num_boxes'])[0:3]
all_box_s = select_boxes(rois=box_info_s[0],
cls_prob=box_info_s[1],
bbox_pred=box_info_s[2],
im_infos=batch_triple['source_im_info'][:,
3:])
all_box_t = select_boxes(rois=box_info_t[0],
cls_prob=box_info_t[1],
bbox_pred=box_info_t[2],
im_infos=batch_triple['target_im_info'][:,
3:])
all_box_r = select_boxes(rois=box_info_r[0],
cls_prob=box_info_r[1],
bbox_pred=box_info_r[2],
im_infos=batch_triple['source_im_info'][:,
3:])
box_s, box_t, box_r = select_box(all_box_s, all_box_t, all_box_r)
theta_st = aff_theta(boxes_s=box_s, boxes_t=box_t)
theta_tr = aff_theta(boxes_s=box_t, boxes_t=box_r)
# theta.shape: (batch_size, 18) for tps
batch_st = {
'source_image': batch_triple['source_image'],
'target_image': batch_triple['target_image']
}
batch_tr = {
'source_image': batch_triple['target_image'],
'target_image': batch_triple['refer_image']
}
theta_aff_tps_st, theta_aff_st = model(
batch_st, theta_st) # from source image to target image
theta_aff_tps_tr, theta_aff_tr = model(
batch_tr, theta_tr) # from target image to refer image
# show_images(batch_st=batch_st, batch_tr=batch_tr, box_s=box_s, box_t=box_t, box_r=box_r)
loss = loss_fn(theta_st=theta_aff_tps_st,
theta_tr=theta_aff_tps_tr,
theta_GT=batch_triple['theta_GT'])
loss.backward()
optimizer.step()
epoch_loss += loss.item()
if (batch_idx + 1) % log_interval == 0:
end = time.time()
print(
'Train epoch: {} [{}/{} ({:.0%})]\t\tCurrent batch loss: {:.6f}\t\tTime cost ({} batches): {:.4f} s'
.format(epoch, batch_idx + 1,
len(dataloader), (batch_idx + 1) / len(dataloader),
loss.item(), batch_idx + 1, end - begin))
if (epoch % 5 == 0 or epoch == 1) and vis is not None:
if (batch_idx + 1) % stride_images == 0 or batch_idx == 0:
watch_images = add_watch(
watch_images, batch_st, batch_tr, tpsTnf, theta_aff_tps_st,
theta_aff_tps_tr,
int((batch_idx + 1) / stride_images) * 6)
if (batch_idx + 1) % stride_loss == 0 or batch_idx == 0:
iter_loss[int((batch_idx + 1) /
stride_loss)] = epoch_loss / (batch_idx + 1)
# tmp_images = batch_triple['target_im'].permute(0, 2, 3, 1) + pixel_means
# tmp_images = tmp_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
# vis.image(torchvision.utils.make_grid(tmp_images, nrow=1, padding=3))
if show:
# if dual:
# warped_image_aff = affTnf(batch_st['source_image'], theta_aff_st_1)
# warped_image_aff_2 = affTnf(batch_tr['source_image'], theta_aff_tr_1)
# warped_image_aff_3 = affTnf(warped_image_aff, theta_aff_tr_1)
# show_images(batch_st, batch_tr, warped_image_aff.detach(), warped_image_aff_2.detach(), warped_image_aff_3.detach())
#
# warped_image_aff = affTnf(batch_st['source_image'], theta_aff_st)
# warped_image_aff_2 = affTnf(batch_tr['source_image'], theta_aff_tr)
# warped_image_aff_3 = affTnf(warped_image_aff, theta_aff_tr)
# show_images(batch_st, batch_tr, warped_image_aff.detach(), warped_image_aff_2.detach(), warped_image_aff_3.detach())
#
# warped_image_aff_tps = tpsTnf(batch_st['source_image'], theta_aff_tps_st)
# warped_image_aff_tps_2 = tpsTnf(batch_tr['source_image'], theta_aff_tps_tr)
# warped_image_aff_tps_3 = tpsTnf(warped_image_aff_tps, theta_aff_tps_tr)
# show_images(batch_st, batch_tr, warped_image_aff_tps.detach(), warped_image_aff_tps_2.detach(), warped_image_aff_tps_3.detach())
#
# warped_image = affTnf(batch_st['source_image'], theta_aff_st_1)
# warped_image = affTnf(warped_image, theta_aff_st)
# warped_image = tpsTnf(warped_image, theta_aff_tps_st)
# warped_image_2 = affTnf(batch_tr['source_image'], theta_aff_tr_1)
# warped_image_2 = affTnf(warped_image_2, theta_aff_tr)
# warped_image_2 = tpsTnf(warped_image_2, theta_aff_tps_tr)
# warped_image_3 = affTnf(warped_image, theta_aff_tr_1)
# warped_image_3 = affTnf(warped_image_3, theta_aff_tr)
# warped_image_3 = tpsTnf(warped_image_3, theta_aff_tps_tr)
# show_images(batch_st, batch_tr, warped_image.detach(), warped_image_2.detach(), warped_image_3.detach())
# else:
# warped_image_aff = affTnf(batch_st['source_image'], theta_st)
# warped_image_aff_2 = affTnf(batch_tr['source_image'], theta_tr)
# warped_image_aff_3 = affTnf(warped_image_aff, theta_tr)
# show_images(batch_st, batch_tr, warped_image_aff.detach(), warped_image_aff_2.detach(), warped_image_aff_3.detach())
warped_image_tps = tpsTnf(batch_st['source_image'],
theta_aff_tps_st)
warped_image_tps_2 = tpsTnf(batch_tr['source_image'],
theta_aff_tps_tr)
warped_image_tps_3 = tpsTnf(warped_image_tps, theta_aff_tps_tr)
show_images(batch_triple, warped_image_tps.detach(),
warped_image_tps_2.detach(),
warped_image_tps_3.detach(), box_s, box_t, box_r)
end = time.time()
# Visualize watch images & train loss
if (epoch % 5 == 0 or epoch == 1) and vis is not None:
opts = dict(jpgquality=100,
title='Epoch ' + str(epoch) +
' source warped_sr target warped_tr refer warped_sr')
# Un-normalize for caffe resnet and vgg
# watch_images = watch_images.permute(0, 2, 3, 1) + pixel_means
# watch_images = watch_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
watch_images = normalize_image(watch_images, forward=False) * 255.0
vis.image(torchvision.utils.make_grid(watch_images, nrow=6, padding=3),
opts=opts)
# vis.images(watch_images, nrow=6, padding=3, opts=opts)
opts_loss = dict(
xlabel='Iterations (' + str(stride_loss) + ')',
ylabel='Loss',
title='GM ResNet101 Detect&Affine&TPS Training Loss in Epoch ' +
str(epoch),
legend=['Loss'],
width=2000)
vis.line(iter_loss, np.arange(36), opts=opts_loss)
epoch_loss /= len(dataloader)
print('Train set -- Average loss: {:.6f}\t\tTime cost: {:.4f}'.format(
epoch_loss, end - begin))
return epoch_loss, end - begin
def vis_control2(vis,
dataloader,
theta_1,
theta_2,
dataset_name,
use_cuda=True):
# Visualize watch images
tpsTnf = GeometricTnf2(geometric_model='tps', use_cuda=use_cuda)
group_size = 5
watch_images = torch.ones(len(dataloader) * group_size, 3, 340, 340)
if use_cuda:
watch_images = watch_images.cuda()
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32))
for batch_idx, batch in enumerate(dataloader):
if use_cuda:
batch = batch_cuda(batch)
# Theta and theta_inver
theta_tps_1 = theta_1['tps'][batch_idx].unsqueeze(0)
theta_tps_2 = theta_2['tps'][batch_idx].unsqueeze(0)
# Warped image
warped_tps_1 = tpsTnf(batch['source_image'], theta_tps_1)
warped_tps_2 = tpsTnf(batch['source_image'], theta_tps_2)
watch_images[batch_idx * group_size, :, 50:290,
50:290] = batch['source_image']
watch_images[batch_idx * group_size + 1, :, 50:290,
50:290] = warped_tps_1
watch_images[batch_idx * group_size + 2, :, 50:290,
50:290] = batch['source_image']
watch_images[batch_idx * group_size + 3, :, 50:290,
50:290] = warped_tps_2
watch_images[batch_idx * group_size + 4, :, 50:290,
50:290] = batch['target_image']
opts = dict(jpgquality=100, title=dataset_name)
watch_images[:, :, 50:290,
50:290] = normalize_image(watch_images[:, :, 50:290, 50:290],
forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3,
1).cpu().numpy().astype(np.uint8)
im_size = torch.Tensor([[240, 240]]).cuda()
for i in range(watch_images.shape[0]):
if i % group_size < 4:
if i % group_size == 0:
cp_norm = theta_1['tps'][int(i / group_size)][:18].view(
1, 2, -1)
if i % group_size == 1:
cp_norm = theta_1['tps'][int(i / group_size)][18:].view(
1, 2, -1)
if i % group_size == 2:
cp_norm = theta_2['tps'][int(i / group_size)][:18].view(
1, 2, -1)
if i % group_size == 3:
cp_norm = theta_2['tps'][int(i / group_size)][18:].view(
1, 2, -1)
watch_images[i] = draw_grid(watch_images[i], cp_norm)
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images, nrow=5, padding=5),
opts=opts)
def vis_pf(vis,
dataloader,
theta_1,
theta_2,
theta_inver_1,
theta_inver_2,
results_1,
results_2,
dataset_name,
use_cuda=True):
# Visualize watch images
tpsTnf_1 = GeometricTnf(geometric_model='tps', use_cuda=use_cuda)
tpsTnf_2 = GeometricTnf2(geometric_model='tps', use_cuda=use_cuda)
pt_1 = PointTnf(use_cuda=use_cuda)
pt_2 = PointTPS(use_cuda=use_cuda)
group_size = 4
watch_images = torch.ones(len(dataloader) * group_size, 3, 280, 240)
watch_keypoints = -torch.ones(len(dataloader) * group_size, 2, 20)
if use_cuda:
watch_images = watch_images.cuda()
watch_keypoints = watch_keypoints.cuda()
num_points = np.ones(len(dataloader) * 6).astype(np.int8)
correct_index = list()
image_names = list()
metrics = list()
# Colors for keypoints
cmap = plt.get_cmap('tab20')
colors = list()
for c in range(20):
r = cmap(c)[0] * 255
g = cmap(c)[1] * 255
b = cmap(c)[2] * 255
colors.append((b, g, r))
fnt = cv2.FONT_HERSHEY_COMPLEX
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32))
for batch_idx, batch in enumerate(dataloader):
if use_cuda:
batch = batch_cuda(batch)
# Theta and theta_inver
theta_tps_1 = theta_1['tps'][batch_idx].unsqueeze(0)
theta_tps_2 = theta_2['tps'][batch_idx].unsqueeze(0)
thetai_tps_1 = theta_inver_1['tps'][batch_idx].unsqueeze(0)
thetai_tps_2 = theta_inver_2['tps'][batch_idx].unsqueeze(0)
# Warped image
warped_tps_1 = tpsTnf_1(batch['source_image'], theta_tps_1)
warped_tps_2 = tpsTnf_2(batch['source_image'], theta_tps_2)
watch_images[batch_idx * group_size, :,
0:240, :] = batch['source_image']
watch_images[batch_idx * group_size + 1, :, 0:240, :] = warped_tps_1
watch_images[batch_idx * group_size + 2, :, 0:240, :] = warped_tps_2
watch_images[batch_idx * group_size + 3, :,
0:240, :] = batch['target_image']
# Warped keypoints
source_im_size = batch['source_im_info'][:, 0:3]
target_im_size = batch['target_im_info'][:, 0:3]
source_points = batch['source_points']
target_points = batch['target_points']
source_points_norm = PointsToUnitCoords(P=source_points,
im_size=source_im_size)
target_points_norm = PointsToUnitCoords(P=target_points,
im_size=target_im_size)
warped_points_tps_norm_1 = pt_1.tpsPointTnf(theta=thetai_tps_1,
points=source_points_norm)
warped_points_tps_1 = PointsToPixelCoords(P=warped_points_tps_norm_1,
im_size=target_im_size)
pck_tps_1, index_tps_1, N_pts = pck(target_points, warped_points_tps_1,
dataset_name)
warped_points_tps_1 = relocate(warped_points_tps_1, target_im_size)
warped_points_tps_norm_2 = pt_2.tpsPointTnf(theta=thetai_tps_2,
points=source_points_norm)
warped_points_tps_2 = PointsToPixelCoords(P=warped_points_tps_norm_2,
im_size=target_im_size)
pck_tps_2, index_tps_2, _ = pck(target_points, warped_points_tps_2,
dataset_name)
warped_points_tps_2 = relocate(warped_points_tps_2, target_im_size)
watch_keypoints[batch_idx * group_size, :, :N_pts] = relocate(
batch['source_points'], source_im_size)[:, :, :N_pts]
watch_keypoints[batch_idx * group_size +
1, :, :N_pts] = warped_points_tps_1[:, :, :N_pts]
watch_keypoints[batch_idx * group_size +
2, :, :N_pts] = warped_points_tps_2[:, :, :N_pts]
watch_keypoints[batch_idx * group_size + 3, :, :N_pts] = relocate(
batch['target_points'], target_im_size)[:, :, :N_pts]
num_points[batch_idx * group_size:batch_idx * group_size +
group_size] = N_pts
correct_index.append(np.arange(N_pts))
correct_index.append(index_tps_1)
correct_index.append(index_tps_2)
correct_index.append(np.arange(N_pts))
image_names.append('Source')
image_names.append('TPS')
image_names.append('TPS_Jitter')
image_names.append('Target')
metrics.append('')
metrics.append('PCK: {:.2%}'.format(pck_tps_1))
metrics.append('PCK: {:.2%}'.format(pck_tps_2))
metrics.append('')
opts = dict(jpgquality=100, title=dataset_name)
# Un-normalize for caffe resnet and vgg
# watch_images = watch_images.permute(0, 2, 3, 1) + pixel_means
# watch_images = watch_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
# watch_images = normalize_image(watch_images, forward=False) * 255.0
watch_images[:, :, 0:240, :] = normalize_image(watch_images[:, :,
0:240, :],
forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3,
1).cpu().numpy().astype(np.uint8)
watch_keypoints = watch_keypoints.cpu().numpy()
for i in range(watch_images.shape[0]):
pos_name = (80, 255)
if (i + 1) % group_size == 1 or (i + 1) % group_size == 0:
pos_pck = (0, 0)
else:
pos_pck = (70, 275)
cv2.putText(watch_images[i], image_names[i], pos_name, fnt, 0.5,
(0, 0, 0), 1)
cv2.putText(watch_images[i], metrics[i], pos_pck, fnt, 0.5, (0, 0, 0),
1)
if (i + 1) % group_size == 0:
for j in range(num_points[i]):
cv2.drawMarker(
watch_images[i],
(watch_keypoints[i, 0, j], watch_keypoints[i, 1, j]),
colors[j], cv2.MARKER_DIAMOND, 12, 2, cv2.LINE_AA)
else:
for j in correct_index[i]:
cv2.drawMarker(
watch_images[i],
(watch_keypoints[i, 0, j], watch_keypoints[i, 1, j]),
colors[j], cv2.MARKER_CROSS, 12, 2, cv2.LINE_AA)
cv2.drawMarker(watch_images[i],
(watch_keypoints[i + (group_size - 1) -
(i % group_size), 0, j],
watch_keypoints[i + (group_size - 1) -
(i % group_size), 1, j]),
colors[j], cv2.MARKER_DIAMOND, 12, 2,
cv2.LINE_AA)
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images, nrow=4, padding=5),
opts=opts)
def vis_control(vis,
dataloader,
theta_1,
theta_2,
dataset_name,
use_cuda=True):
# Visualize watch images
tpsTnf_1 = GeometricTnf(geometric_model='tps', use_cuda=use_cuda)
tpsTnf_2 = GeometricTnf2(geometric_model='tps', use_cuda=use_cuda)
group_size = 5
watch_images = torch.ones(len(dataloader) * group_size, 3, 340, 340)
if use_cuda:
watch_images = watch_images.cuda()
# Colors for keypoints
cmap = plt.get_cmap('tab20')
colors = list()
for c in range(20):
r = cmap(c)[0] * 255
g = cmap(c)[1] * 255
b = cmap(c)[2] * 255
colors.append((b, g, r))
fnt = cv2.FONT_HERSHEY_COMPLEX
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32))
for batch_idx, batch in enumerate(dataloader):
if use_cuda:
batch = batch_cuda(batch)
# Theta and theta_inver
theta_tps_1 = theta_1['tps'][batch_idx].unsqueeze(0)
theta_tps_2 = theta_2['tps'][batch_idx].unsqueeze(0)
# Warped image
warped_tps_1 = tpsTnf_1(batch['source_image'], theta_tps_1)
warped_tps_2 = tpsTnf_2(batch['source_image'], theta_tps_2)
watch_images[batch_idx * group_size, :, 50:290,
50:290] = batch['source_image']
watch_images[batch_idx * group_size + 1, :, 50:290,
50:290] = warped_tps_1
watch_images[batch_idx * group_size + 2, :, 50:290,
50:290] = batch['source_image']
watch_images[batch_idx * group_size + 3, :, 50:290,
50:290] = warped_tps_2
watch_images[batch_idx * group_size + 4, :, 50:290,
50:290] = batch['target_image']
opts = dict(jpgquality=100, title=dataset_name)
watch_images[:, :, 50:290,
50:290] = normalize_image(watch_images[:, :, 50:290, 50:290],
forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3,
1).cpu().numpy().astype(np.uint8)
im_size = torch.Tensor([[240, 240]]).cuda()
for i in range(watch_images.shape[0]):
if i % group_size == 0:
cp_norm = theta_1['tps'][int(i / group_size)].view(1, 2, -1)
cp = PointsToPixelCoords(P=cp_norm, im_size=im_size)
cp = cp.squeeze().cpu().numpy() + 50
for j in range(9):
cv2.drawMarker(watch_images[i], (cp[0, j], cp[1, j]),
(0, 0, 255), cv2.MARKER_TILTED_CROSS, 12, 2,
cv2.LINE_AA)
for j in range(2):
for k in range(3):
# vertical grid
cv2.line(watch_images[i],
(cp[0, j + k * 3], cp[1, j + k * 3]),
(cp[0, j + k * 3 + 1], cp[1, j + k * 3 + 1]),
(0, 0, 255), 2, cv2.LINE_AA)
# horizontal grid
cv2.line(watch_images[i],
(cp[0, j * 3 + k], cp[1, j * 3 + k]),
(cp[0, j * 3 + k + 3], cp[1, j * 3 + k + 3]),
(0, 0, 255), 2, cv2.LINE_AA)
if i % group_size == 1:
cp_norm = torch.Tensor(
[-1, -1, -1, 0, 0, 0, 1, 1, 1, -1, 0, 1, -1, 0, 1, -1, 0,
1]).cuda().view(1, 2, -1)
cp = PointsToPixelCoords(P=cp_norm, im_size=im_size)
cp = cp.squeeze().cpu().numpy() + 50
for j in range(9):
cv2.drawMarker(watch_images[i], (cp[0, j], cp[1, j]),
(0, 0, 255), cv2.MARKER_TILTED_CROSS, 12, 2,
cv2.LINE_AA)
for j in range(1):
for k in range(3):
# vertical grid
cv2.line(watch_images[i],
(cp[0, j + k * 3], cp[1, j + k * 3]),
(cp[0, j + k * 3 + 1], cp[1, j + k * 3 + 1]),
(0, 0, 255), 2, cv2.LINE_AA)
# horizontal grid
cv2.line(watch_images[i],
(cp[0, j * 3 + k], cp[1, j * 3 + k]),
(cp[0, j * 3 + k + 3], cp[1, j * 3 + k + 3]),
(0, 0, 255), 2, cv2.LINE_AA)
if i % group_size == 2:
cp_norm = theta_2['tps'][int(i / group_size)][:18].view(1, 2, -1)
cp = PointsToPixelCoords(P=cp_norm, im_size=im_size)
cp = cp.squeeze().cpu().numpy() + 50
for j in range(9):
cv2.drawMarker(watch_images[i], (cp[0, j], cp[1, j]),
(0, 0, 255), cv2.MARKER_TILTED_CROSS, 12, 2,
cv2.LINE_AA)
for j in range(2):
for k in range(3):
# vertical grid
cv2.line(watch_images[i],
(cp[0, j + k * 3], cp[1, j + k * 3]),
(cp[0, j + k * 3 + 1], cp[1, j + k * 3 + 1]),
(0, 0, 255), 2, cv2.LINE_AA)
# horizontal grid
cv2.line(watch_images[i],
(cp[0, j * 3 + k], cp[1, j * 3 + k]),
(cp[0, j * 3 + k + 3], cp[1, j * 3 + k + 3]),
(0, 0, 255), 2, cv2.LINE_AA)
if i % group_size == 3:
cp_norm = theta_2['tps'][int(i / group_size)][18:].view(1, 2, -1)
cp = PointsToPixelCoords(P=cp_norm, im_size=im_size)
cp = cp.squeeze().cpu().numpy() + 50
for j in range(9):
cv2.drawMarker(watch_images[i], (cp[0, j], cp[1, j]),
(0, 0, 255), cv2.MARKER_TILTED_CROSS, 12, 2,
cv2.LINE_AA)
for j in range(2):
for k in range(3):
# vertical grid
cv2.line(watch_images[i],
(cp[0, j + k * 3], cp[1, j + k * 3]),
(cp[0, j + k * 3 + 1], cp[1, j + k * 3 + 1]),
(0, 0, 255), 2, cv2.LINE_AA)
# horizontal grid
cv2.line(watch_images[i],
(cp[0, j * 3 + k], cp[1, j * 3 + k]),
(cp[0, j * 3 + k + 3], cp[1, j * 3 + k + 3]),
(0, 0, 255), 2, cv2.LINE_AA)
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images, nrow=5, padding=5),
opts=opts)
def train_fn_dual(epoch, model, loss_fn, optimizer, dataloader, triple_generation, use_cuda=True, log_interval=100, vis=None, show=False):
"""
Train the model with synthetically training triple:
{source image, target image, refer image (warped source image), theta_GT} from PF-PASCAL.
1. Train the transformation parameters theta_st from source image to target image;
2. Train the transformation parameters theta_tr from target image to refer image;
3. Combine theta_st and theta_st to obtain theta from source image to refer image, and compute loss between
theta and theta_GT.
"""
geoTnf = ComposedGeometricTnf(use_cuda=use_cuda)
affTnf = GeometricTnf(geometric_model='affine', use_cuda=use_cuda)
tpsTnf = GeometricTnf(geometric_model='tps', use_cuda=use_cuda)
epoch_loss = 0
if (epoch % 5 == 0 or epoch == 1) and vis is not None:
stride_images = len(dataloader) / 3
group_size = 6
watch_images = torch.ones(group_size * 4, 3, 260, 240).cuda()
# watch_images = torch.ones(group_size * 20, 3, 260, 240).cuda()
image_names = list()
fnt = cv2.FONT_HERSHEY_COMPLEX
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32)).cuda()
stride_loss = len(dataloader) / 105
iter_loss = np.zeros(106)
begin = time.time()
for batch_idx, batch in enumerate(dataloader):
''' Move input batch to gpu '''
# batch['source_image'].shape & batch['target_image'].shape: (batch_size, 3, 240, 240)
# batch['theta'].shape-tps: (batch_size, 18)-random or (batch_size, 18, 1, 1)-(pre-set from csv)
if use_cuda:
batch = batch_cuda(batch)
''' Get the training triple {source image, target image, refer image (warped source image), theta_GT}'''
batch_triple = triple_generation(batch)
''' Train the model '''
optimizer.zero_grad()
# Predict tps parameters between images
# theta.shape: (batch_size, 18) for tps
batch_st = {'source_image': batch_triple['source_image'], 'target_image': batch_triple['target_image']}
batch_tr = {'source_image': batch_triple['target_image'], 'target_image': batch_triple['refer_image']}
theta_aff_tps_st, theta_aff_st = model(batch_st) # from source image to target image
theta_aff_tps_tr, theta_aff_tr = model(batch_tr) # from target image to refer image
# show_images(batch_st=batch_st, batch_tr=batch_tr, box_s=box_s, box_t=box_t, box_r=box_r)
# loss = loss_fn(theta_st=theta_aff_tps_st, theta_tr=theta_aff_tps_tr, theta_GT=batch_triple['theta_GT'])
loss = loss_fn(theta_aff_tps_st=theta_aff_tps_st, theta_aff_st=theta_aff_st, theta_aff_tps_tr=theta_aff_tps_tr,
theta_aff_tr=theta_aff_tr, theta_aff_GT=batch_triple['theta_aff_GT'], theta_tps_GT=batch_triple['theta_tps_GT'])
loss.backward()
optimizer.step()
epoch_loss += loss.item()
if (batch_idx+1) % log_interval == 0:
end = time.time()
print('Train epoch: {} [{}/{} ({:.0%})]\t\tCurrent batch loss: {:.6f}\t\tTime cost ({} batches): {:.4f} s'
.format(epoch, batch_idx+1, len(dataloader), (batch_idx+1) / len(dataloader), loss.item(), batch_idx + 1, end - begin))
if (epoch % 5 == 0 or epoch == 1) and vis is not None:
if (batch_idx + 1) % stride_images == 0 or batch_idx == 0:
watch_images, image_names = add_watch(watch_images, image_names, batch_st, batch_tr, affTnf, tpsTnf, geoTnf, theta_aff_tps_st, theta_aff_st, theta_aff_tps_tr, theta_aff_tr, int((batch_idx + 1) / stride_images) * group_size)
# if batch_idx <= 19:
# watch_images, image_names = add_watch(watch_images, image_names, batch_st, batch_tr, affTnf, tpsTnf, geoTnf, theta_aff_tps_st, theta_aff_st, theta_aff_tps_tr, theta_aff_tr, batch_idx * group_size, batch_triple)
# if batch_idx == 19:
# opts = dict(jpgquality=100, title='Epoch ' + str(epoch) + ' source warped_st target warped_tr refer warped_sr')
# watch_images[:, :, 0:240, :] = normalize_image(watch_images[:, :, 0:240, :], forward=False)
# watch_images *= 255.0
# watch_images = watch_images.permute(0, 2, 3, 1).cpu().numpy().astype(np.uint8)
# for i in range(len(image_names)):
# cv2.putText(watch_images[i], image_names[i], (80, 255), fnt, 0.5, (0, 0, 0), 1)
# watch_images = torch.Tensor(watch_images.astype(np.float32))
# watch_images = watch_images.permute(0, 3, 1, 2)
# vis.image(torchvision.utils.make_grid(watch_images, nrow=group_size, padding=3), opts=opts)
if (batch_idx + 1) % stride_loss == 0 or batch_idx == 0:
iter_loss[int((batch_idx + 1) / stride_loss)] = epoch_loss / (batch_idx + 1)
# tmp_images = normalize_image(batch_tr['target_image'], forward=False) * 255.0
# vis.images(tmp_images, nrow=8, padding=3)
# if show:
# if dual:
# warped_image_aff = affTnf(batch_st['source_image'], theta_aff_st_1)
# warped_image_aff_2 = affTnf(batch_tr['source_image'], theta_aff_tr_1)
# warped_image_aff_3 = affTnf(warped_image_aff, theta_aff_tr_1)
# show_images(batch_st, batch_tr, warped_image_aff.detach(), warped_image_aff_2.detach(), warped_image_aff_3.detach())
#
# warped_image_aff = affTnf(batch_st['source_image'], theta_aff_st)
# warped_image_aff_2 = affTnf(batch_tr['source_image'], theta_aff_tr)
# warped_image_aff_3 = affTnf(warped_image_aff, theta_aff_tr)
# show_images(batch_st, batch_tr, warped_image_aff.detach(), warped_image_aff_2.detach(), warped_image_aff_3.detach())
#
# warped_image_aff_tps = tpsTnf(batch_st['source_image'], theta_aff_tps_st)
# warped_image_aff_tps_2 = tpsTnf(batch_tr['source_image'], theta_aff_tps_tr)
# warped_image_aff_tps_3 = tpsTnf(warped_image_aff_tps, theta_aff_tps_tr)
# show_images(batch_st, batch_tr, warped_image_aff_tps.detach(), warped_image_aff_tps_2.detach(), warped_image_aff_tps_3.detach())
#
# warped_image = affTnf(batch_st['source_image'], theta_aff_st_1)
# warped_image = affTnf(warped_image, theta_aff_st)
# warped_image = tpsTnf(warped_image, theta_aff_tps_st)
# warped_image_2 = affTnf(batch_tr['source_image'], theta_aff_tr_1)
# warped_image_2 = affTnf(warped_image_2, theta_aff_tr)
# warped_image_2 = tpsTnf(warped_image_2, theta_aff_tps_tr)
# warped_image_3 = affTnf(warped_image, theta_aff_tr_1)
# warped_image_3 = affTnf(warped_image_3, theta_aff_tr)
# warped_image_3 = tpsTnf(warped_image_3, theta_aff_tps_tr)
# show_images(batch_st, batch_tr, warped_image.detach(), warped_image_2.detach(), warped_image_3.detach())
# else:
# warped_image_aff = affTnf(batch_st['source_image'], theta_st)
# warped_image_aff_2 = affTnf(batch_tr['source_image'], theta_tr)
# warped_image_aff_3 = affTnf(warped_image_aff, theta_tr)
# show_images(batch_st, batch_tr, warped_image_aff.detach(), warped_image_aff_2.detach(), warped_image_aff_3.detach())
# warped_image_tps = tpsTnf(batch_st['source_image'], theta_st)
# warped_image_tps_2 = tpsTnf(batch_tr['source_image'], theta_tr)
# warped_image_tps_3 = tpsTnf(warped_image_tps, theta_tr)
# show_images(batch_st, batch_tr, warped_image_tps.detach(), warped_image_tps_2.detach(), warped_image_tps_3.detach())
end = time.time()
# Visualize watch images & train loss
if (epoch % 5 == 0 or epoch == 1) and vis is not None:
opts = dict(jpgquality=100, title='Epoch ' + str(epoch) + ' source warped_st target warped_tr refer warped_sr')
watch_images[:, :, 0:240, :] = normalize_image(watch_images[:, :, 0:240, :], forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3, 1).cpu().numpy().astype(np.uint8)
for i in range(watch_images.shape[0]):
cv2.putText(watch_images[i], image_names[i], (80, 255), fnt, 0.5, (0, 0, 0), 1)
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images, nrow=group_size, padding=3), opts=opts)
# Un-normalize for caffe resnet and vgg
# watch_images = watch_images.permute(0, 2, 3, 1) + pixel_means
# watch_images = watch_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
opts_loss = dict(xlabel='Iterations (' + str(stride_loss) + ')',
ylabel='Loss',
title='GM ResNet101 AffTPS Training Loss in Epoch ' + str(epoch),
legend=['Loss'],
width=2000)
vis.line(iter_loss, np.arange(106), opts=opts_loss)
epoch_loss /= len(dataloader)
print('Train set -- Average loss: {:.6f}\t\tTime cost: {:.4f}'.format(epoch_loss, end - begin))
return epoch_loss, end - begin
def vis_fn_detect(vis,
model,
faster_rcnn,
aff_theta,
train_loss,
val_pck,
train_lr,
epoch,
num_epochs,
dataloader,
use_cuda=True):
# Visualize watch images
affTnf = GeometricTnf(geometric_model='affine', use_cuda=use_cuda)
tpsTnf = GeometricTnf(geometric_model='tps', use_cuda=use_cuda)
watch_images = torch.Tensor(len(dataloader) * 5, 3, 240, 240)
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32))
for batch_idx, batch in enumerate(dataloader):
if use_cuda:
batch = batch_cuda(batch)
box_info_s = faster_rcnn(im_data=batch['source_im'],
im_info=batch['source_im_info'][:, 3:],
gt_boxes=batch['source_gt_boxes'],
num_boxes=batch['source_num_boxes'])[0:3]
box_info_t = faster_rcnn(im_data=batch['target_im'],
im_info=batch['target_im_info'][:, 3:],
gt_boxes=batch['target_gt_boxes'],
num_boxes=batch['target_num_boxes'])[0:3]
all_box_s = select_boxes(rois=box_info_s[0],
cls_prob=box_info_s[1],
bbox_pred=box_info_s[2],
im_infos=batch['source_im_info'][:, 3:])
all_box_t = select_boxes(rois=box_info_t[0],
cls_prob=box_info_t[1],
bbox_pred=box_info_t[2],
im_infos=batch['target_im_info'][:, 3:])
box_s, box_t = select_box_st(all_box_s, all_box_t)
theta_det = aff_theta(boxes_s=box_s, boxes_t=box_t)
theta_aff_tps, theta_aff = model(batch, theta_det)
warped_image_1 = affTnf(batch['source_image'], theta_det)
warped_image_2 = affTnf(warped_image_1, theta_aff)
warped_image_3 = tpsTnf(warped_image_2, theta_aff_tps)
watch_images[batch_idx * 5] = batch['source_image'][0]
watch_images[batch_idx * 5 + 1] = warped_image_1[0]
watch_images[batch_idx * 5 + 2] = warped_image_2[0]
watch_images[batch_idx * 5 + 3] = warped_image_3[0]
watch_images[batch_idx * 5 + 4] = batch['target_image'][0]
opts = dict(jpgquality=100,
title='Epoch ' + str(epoch) + ' source warped target')
# Un-normalize for caffe resnet and vgg
# watch_images = watch_images.permute(0, 2, 3, 1) + pixel_means
# watch_images = watch_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
watch_images = normalize_image(watch_images, forward=False) * 255.0
vis.image(torchvision.utils.make_grid(watch_images, nrow=5, padding=5),
opts=opts)
# vis.images(watch_images, nrow=5, padding=3, opts=opts)
if epoch == num_epochs:
epochs = np.arange(1, num_epochs + 1)
# Visualize train loss
opts_loss = dict(xlabel='Epoch',
ylabel='Loss',
title='GM ResNet101 Detect&Affine&TPS Training Loss',
legend=['Loss'],
width=2000)
vis.line(train_loss, epochs, opts=opts_loss)
# Visualize val pck
opts_pck = dict(xlabel='Epoch',
ylabel='Val PCK',
title='GM ResNet101 Detect&Affine&TPS Val PCK',
legend=['PCK'],
width=2000)
vis.line(val_pck, epochs, opts=opts_pck)
# Visualize train lr
opts_lr = dict(
xlabel='Epoch',
ylabel='Learning Rate',
title='GM ResNet101 Detect&Affine&TPS Training Learning Rate',
legend=['LR'],
width=2000)
vis.line(train_lr, epochs, opts=opts_lr)
def __call__(self, batch=None):
"""
Generate a synthetically training triple:
1. Use the given image pair as the source image and target image;
2. Padding the source image, and wrap the padding image with the given transformation to generate the refer image;
3. The training triple consists of {source image, target image, refer image, theta_GT}
4. The input for fasterRCNN consists of {source_im, target_im, refer_im, source_im_info, target_im_info,
refer_im_info, source_gt_boxes, target_gt_boxes, refer_gt_boxes, source_num_boxes, target_num_boxes,
refer_num_boxes}.
"""
# image_batch.shape: (batch_size, 3, H, W)
# theta_batch.shape-tps: (batch_size, 18)-random or (batch_size, 18, 1, 1)-(pre-set from csv)
# boxes.shape: (batch_size, 4), 4: (x_min, y_min, x_max, y_max)
img_A_batch = batch['source_image']
img_B_batch = batch['target_image']
theta_aff = batch['theta_GT'][:, :6].contiguous()
theta_tps = batch['theta_GT'][:, 6:]
# Generate symmetrically padded image for bigger sampling region to warp the source image
padded_image_batch = self.symmetricImagePad(
image_batch=img_A_batch, padding_factor=self.padding_factor)
img_A_batch.requires_grad = False
img_B_batch.requires_grad = False
padded_image_batch.requires_grad = False
theta_aff.requires_grad = False
theta_tps.requires_grad = False
# Get the refer image by warping the padded image with the given transformation
# warped_image_batch = self.affTnf(image_batch=padded_image_batch, theta_batch=theta_aff, padding_factor=0.5, crop_factor=self.crop_factor)
# warped_image_batch = self.symmetricImagePad(image_batch=warped_image_batch, padding_factor=self.padding_factor)
# warped_image_batch = self.tpsTnf(image_batch=warped_image_batch, theta_batch=theta_tps, padding_factor=0.5, crop_factor=self.crop_factor)
# warped_image_aff = self.affTnf(image_batch=padded_image_batch, theta_batch=theta_aff, padding_factor=self.padding_factor, crop_factor=self.crop_factor)
# warped_image_tps = self.tpsTnf(image_batch=padded_image_batch, theta_batch=theta_tps, padding_factor=self.padding_factor, crop_factor=self.crop_factor)
# warped_image_batch = self.affTpsTnf(source_image=padded_image_batch, theta_aff=theta_aff, theta_aff_tps=theta_tps, use_cuda=self.use_cuda)
warped_image_batch = self.geometricTnf(image_batch=padded_image_batch,
theta_aff=theta_aff,
theta_aff_tps=theta_tps)
# Get the refer im for extracting rois
tmp_image_batch = normalize_image(image=warped_image_batch,
forward=False) * 255.0
tmp_image_batch = tmp_image_batch.cpu().numpy().transpose((0, 2, 3, 1))
tmp_image_batch = tmp_image_batch[:, :, :, ::-1] # RGB -> BGR
warped_im_batch = torch.zeros_like(warped_image_batch,
dtype=torch.float32)
for i in range(tmp_image_batch.shape[0]):
warped_im_batch[i] = roi_data(tmp_image_batch[i])[0]
warped_im_batch = warped_im_batch.cuda()
warped_im_batch.requires_grad = False
# img_A_batch.shape, img_B_batch.shape, warped_image_batch.shape: (batch_size, 3, 240, 240)
# theta_batch.shape-tps: (batch_size, 18)-random or (batch_size, 18, 1, 1)-(pre-set from csv)
# theta_batch.shape-affine: (batch_size, 2, 3)
# return {'source_image': img_A_batch, 'target_image': img_B_batch, 'refer_image': warped_image_batch,
# 'source_im': batch['source_im'], 'target_im': batch['target_im'], 'refer_im': warped_im_batch,
# 'source_im_info': batch['source_im_info'], 'target_im_info': batch['target_im_info'], 'refer_im_info': batch['source_im_info'],
# 'source_gt_boxes': batch['source_gt_boxes'], 'target_gt_boxes': batch['target_gt_boxes'], 'refer_gt_boxes': batch['source_gt_boxes'],
# 'source_num_boxes': batch['source_num_boxes'], 'target_num_boxes': batch['target_num_boxes'], 'refer_num_boxes': batch['source_num_boxes'],
# 'theta_aff_GT': theta_aff, 'theta_tps_GT': theta_tps}
return {
'source_image': img_A_batch,
'target_image': img_B_batch,
'refer_image': warped_image_batch,
'source_im_info': batch['source_im_info'],
'target_im_info': batch['target_im_info'],
'refer_im_info': batch['source_im_info'],
'theta_aff_GT': theta_aff,
'theta_tps_GT': theta_tps
}
def vis_tss(vis,
dataloader,
theta,
theta_weak,
csv_file,
title,
use_cuda=True):
# Visualize watch images
dataframe = pd.read_csv(csv_file)
scores_det = dataframe.iloc[:, 5]
scores_det_aff = dataframe.iloc[:, 6]
scores_det_aff_tps = dataframe.iloc[:, 7]
scores_aff = dataframe.iloc[:, 8]
scores_aff_tps = dataframe.iloc[:, 9]
affTnf = GeometricTnf(geometric_model='affine', use_cuda=use_cuda)
tpsTnf = GeometricTnf(geometric_model='tps', use_cuda=use_cuda)
watch_images = torch.ones(len(dataloader) * 8, 3, 280, 240)
if use_cuda:
watch_images = watch_images.cuda()
image_names = list()
flow = list()
fnt = cv2.FONT_HERSHEY_COMPLEX
# means for normalize of caffe resnet and vgg
# pixel_means = torch.Tensor(np.array([[[[102.9801, 115.9465, 122.7717]]]]).astype(np.float32))
for batch_idx, batch in enumerate(dataloader):
if use_cuda:
batch = batch_cuda(batch)
theta_det = theta['det'][batch_idx].unsqueeze(0)
theta_det_aff = theta['det_aff'][batch_idx].unsqueeze(0)
theta_det_aff_tps = theta['det_aff_tps'][batch_idx].unsqueeze(0)
theta_aff = theta_weak['aff'][batch_idx].unsqueeze(0)
theta_aff_tps = theta_weak['aff_tps'][batch_idx].unsqueeze(0)
# Warped image
warped_det = affTnf(batch['source_image'], theta_det)
warped_det_aff = affTnf(warped_det, theta_det_aff)
warped_det_aff_tps = tpsTnf(warped_det_aff, theta_det_aff_tps)
warped_aff = affTnf(batch['source_image'], theta_aff)
warped_aff_tps = tpsTnf(warped_aff, theta_aff_tps)
watch_images[batch_idx * 8, :, 0:240, :] = batch['source_image']
watch_images[batch_idx * 8 + 1, :, 0:240, :] = warped_det
watch_images[batch_idx * 8 + 2, :, 0:240, :] = warped_det_aff
watch_images[batch_idx * 8 + 3, :, 0:240, :] = warped_det_aff_tps
watch_images[batch_idx * 8 + 4, :, 0:240, :] = batch['target_image']
watch_images[batch_idx * 8 + 6, :, 0:240, :] = warped_aff
watch_images[batch_idx * 8 + 7, :, 0:240, :] = warped_aff_tps
image_names.append('Source')
image_names.append('Det')
image_names.append('Det_aff')
image_names.append('Det_aff_tps')
image_names.append('Target')
image_names.append('')
image_names.append('Rocco_aff')
image_names.append('Rocco_aff_tps')
flow.append('')
flow.append('Flow: {:.3f}'.format(scores_det[batch_idx]))
flow.append('Flow: {:.3f}'.format(scores_det_aff[batch_idx]))
flow.append('Flow: {:.3f}'.format(scores_det_aff_tps[batch_idx]))
flow.append('')
flow.append('')
flow.append('Flow: {:.3f}'.format(scores_aff[batch_idx]))
flow.append('Flow: {:.3f}'.format(scores_aff_tps[batch_idx]))
opts = dict(jpgquality=100, title=title)
# Un-normalize for caffe resnet and vgg
# watch_images = watch_images.permute(0, 2, 3, 1) + pixel_means
# watch_images = watch_images[:, :, :, [2, 1, 0]].permute(0, 3, 1, 2)
# watch_images = normalize_image(watch_images, forward=False) * 255.0
watch_images[:, :, 0:240, :] = normalize_image(watch_images[:, :,
0:240, :],
forward=False)
watch_images *= 255.0
watch_images = watch_images.permute(0, 2, 3,
1).cpu().numpy().astype(np.uint8)
for i in range(watch_images.shape[0]):
if (i + 1) % 8 != 6:
pos_name = (80, 255)
if (i + 1) % 8 == 1 or (i + 1) % 8 == 5:
pos_lt_ac = (0, 0)
pos_flow = (0, 0)
else:
pos_flow = (70, 275)
cv2.putText(watch_images[i], image_names[i], pos_name, fnt, 0.5,
(0, 0, 0), 1)
cv2.putText(watch_images[i], flow[i], pos_flow, fnt, 0.5,
(0, 0, 0), 1)
else:
watch_images[i] = 255
watch_images = torch.Tensor(watch_images.astype(np.float32))
watch_images = watch_images.permute(0, 3, 1, 2)
vis.image(torchvision.utils.make_grid(watch_images, nrow=4, padding=5),
opts=opts)
