Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Adjusting the 300WLP position maps generation #60

Open
wants to merge 5 commits into
base: master
Choose a base branch
from
Open

Adjusting the 300WLP position maps generation #60

Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
ba08ae0
Add files via upload
Nnemr Feb 19, 2019
eb522ee
Add files via upload
Nnemr Feb 20, 2019
883d0c9
Add files via upload
Nnemr Feb 21, 2019
b86b834
Add files via upload
Nnemr Feb 21, 2019
9994277
Update 8_generate_posmap_300WLP.py
Nnemr Oct 4, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
117 changes: 76 additions & 41 deletions examples/8_generate_posmap_300WLP.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,11 +4,13 @@
import os, sys
import numpy as np
import scipy.io as sio
from skimage import io
import random as ran
from skimage.transform import SimilarityTransform
from skimage import io, util
import skimage.transform
from time import time
import cv2
import matplotlib.pyplot as plt

sys.path.append('..')
import face3d
from face3d import mesh
Expand All @@ -24,21 +26,20 @@ def process_uv(uv_coords, uv_h = 256, uv_w = 256):
def run_posmap_300W_LP(bfm, image_path, mat_path, save_folder, uv_h = 256, uv_w = 256, image_h = 256, image_w = 256):
# 1. load image and fitted parameters
image_name = image_path.strip().split('/')[-1]
image = io.imread(image_path)/255.
[h, w, c] = image.shape

info = sio.loadmat(mat_path)
pose_para = info['Pose_Para'].T.astype(np.float32)
shape_para = info['Shape_Para'].astype(np.float32)
exp_para = info['Exp_Para'].astype(np.float32)

# 2. generate mesh
image = io.imread(image_path)/255;
[h, w, c] = image.shape;

info = sio.loadmat(mat_path);
pose_para = info['Pose_Para'].T.astype(np.float32);
shape_para = info['Shape_Para'].astype(np.float32);
exp_para = info['Exp_Para'].astype(np.float32);
# 2. generate mesh;
# generate shape
vertices = bfm.generate_vertices(shape_para, exp_para)
vertices = bfm.generate_vertices(shape_para, exp_para);
# transform mesh
s = pose_para[-1, 0]
angles = pose_para[:3, 0]
t = pose_para[3:6, 0]
s = pose_para[-1, 0];
angles = pose_para[:3, 0];
t = pose_para[3:6, 0];
transformed_vertices = bfm.transform_3ddfa(vertices, s, angles, t)
projected_vertices = transformed_vertices.copy() # using stantard camera & orth projection as in 3DDFA
image_vertices = projected_vertices.copy()
Expand All @@ -58,55 +59,89 @@ def run_posmap_300W_LP(bfm, image_path, mat_path, save_folder, uv_h = 256, uv_w
marg = old_size*0.1
t_x = np.random.rand()*marg*2 - marg
t_y = np.random.rand()*marg*2 - marg
center[0] = center[0]+t_x; center[1] = center[1]+t_y
center[0] = center[0]+t_x;
center[1] = center[1]+t_y
size = size*(np.random.rand()*0.2 + 0.9)

# crop and record the transform parameters
src_pts = np.array([[center[0]-size/2, center[1]-size/2], [center[0] - size/2, center[1]+size/2], [center[0]+size/2, center[1]-size/2]])
DST_PTS = np.array([[0, 0], [0, image_h - 1], [image_w - 1, 0]])
tform = skimage.transform.estimate_transform('similarity', src_pts, DST_PTS)
cropped_image = skimage.transform.warp(image, tform.inverse, output_shape=(image_h, image_w))

src_pts = np.array([[center[0]-size/2, center[1]-size/2], [center[0] - size/2, center[1]+size/2], [center[0]+size/2, center[1]-size/2]]);
DST_PTS = np.array([[0, 0], [0, image_h - 1], [image_w - 1, 0]]);
tform = skimage.transform.estimate_transform('similarity', src_pts, DST_PTS);
# transform face position(image vertices) along with 2d facial image
angle = np.random.rand() * 90 - 45;
rows, cols = image.shape[0], image.shape[1];
# rotation around center
center = np.array((cols, rows)) / 2. - 0.5;
#Augmenting the dataset
tform1 = SimilarityTransform(translation=center);
tform2 = SimilarityTransform(rotation=np.deg2rad(angle));
tform3 = SimilarityTransform(translation=-center);
rotate_transform = tform3 + tform2 + tform1;
tform = rotate_transform + tform;
#end of augmentation
opt = ran.randint(1,2);
cropped_image = skimage.transform.warp(image, tform.inverse, output_shape=(image_h, image_w));
position = image_vertices.copy()
position[:, 2] = 1
position = np.dot(position, tform.params.T)
position[:, 2] = image_vertices[:, 2]*tform.params[0, 0] # scale z
position[:, 2] = position[:, 2] - np.min(position[:, 2]) # translate z

# 4. uv position map: render position in uv space
uv_position_map = mesh.render.render_colors(uv_coords, bfm.full_triangles, position, uv_h, uv_w, c = 3)

#cv2.imshow('image', cropped_image);
#cv2.waitKey(0);
#cv2.destroyAllWindows();
# 5. save files
io.imsave('{}/{}'.format(save_folder, image_name), np.squeeze(cropped_image))
np.save('{}/{}'.format(save_folder, image_name.replace('jpg', 'npy')), uv_position_map)
io.imsave('{}/{}'.format(save_folder, image_name.replace('.jpg', '_posmap.jpg')), (uv_position_map)/max(image_h, image_w)) # only for show

io.imsave('{}\{}'.format(save_folder, image_name), np.squeeze(cropped_image));
np.save('{}\{}'.format(save_folder, image_name.replace('jpg', 'npy')), uv_position_map);
io.imsave('{}\{}'.format('results/uv_maps/', image_name.replace('.jpg', '_posmap.jpg')), (uv_position_map)/abs(uv_position_map.max())); # only for show
#cv2.imwrite(image_name[:-4]+'_posmap.jpg',uv_position_map);
# --verify
# import cv2
# uv_texture_map_rec = cv2.remap(cropped_image, uv_position_map[:,:,:2].astype(np.float32), None, interpolation=cv2.INTER_LINEAR, borderMode=cv2.BORDER_CONSTANT,borderValue=(0))
# io.imsave('{}/{}'.format(save_folder, image_name.replace('.jpg', '_tex.jpg')), np.squeeze(uv_texture_map_rec))
#import cv2
uv_texture_map_rec = cv2.remap(cropped_image, uv_position_map[:,:,:2].astype(np.float32), None, interpolation=cv2.INTER_LINEAR,borderMode= cv2.BORDER_CONSTANT,borderValue=(0));
#io.imsave('{}\{}'.format(save_folder, image_name.replace('.jpg', '_tex.jpg')), np.squeeze(uv_texture_map_rec)); #Save fitted face on position map (texture).


if __name__ == '__main__':
save_folder = 'results/posmap_300WLP'
save_folder = 'results/'
if not os.path.exists(save_folder):
os.mkdir(save_folder)

# set para
uv_h = uv_w = 256
image_h = image_w = 256

image_h = image_w = 256;
# load uv coords
global uv_coords
uv_coords = face3d.morphable_model.load.load_uv_coords('Data/BFM/Out/BFM_UV.mat') #
uv_coords = face3d.morphable_model.load.load_uv_coords('Data/BFM/Out/BFM_UV.mat')
uv_coords = process_uv(uv_coords, uv_h, uv_w)

'''
Save LFPW Testing as well, only the first 8.
'''
# load bfm
bfm = MorphabelModel('Data/BFM/Out/BFM.mat')

# run
image_path = 'Data/IBUG_image_008_1_0.jpg'
mat_path = 'Data/IBUG_image_008_1_0.mat'
run_posmap_300W_LP(bfm, image_path, mat_path, save_folder)

content = [];
print('Running');
s =0;
f=0;
types = ['AFW', 'HELEN', 'LFPW', 'IBUG', 'LFPW_Test'];
for i in types:
print(i);
with open(i+'_Data.txt', 'r') as fileRead:
content = [file.rstrip('\n') for file in fileRead];
s=0;
print(len(content));
for filename in content:
#print(filename)
#if(s==8 and i is 'LFPW_Test'):
# break
if(s%500 ==0):
print(str(s) +'/' +str(len(content)))
image_path = 'Data/BFM/300W_LP/'+ i+'/'+filename+'.jpg';
mat_path = 'Data/BFM/300W_LP/'+i+'/'+filename+'.mat';
if(i is 'LFPW_Test'):
image_path = 'Data/BFM/300W_LP/'+ 'LFPW'+'/'+filename+'.jpg';
mat_path = 'Data/BFM/300W_LP/'+'LFPW'+'/'+filename+'.mat';
run_posmap_300W_LP(bfm, image_path, mat_path, save_folder)
s+=1;
print(s+f)
print(f)