def __getitem__(self, index):
image_path = self.images[index]
kps = self.kp2ds[index].copy()
box = self.boxs[index]
scale = np.random.rand(4) * (self.scale_range[1] -
self.scale_range[0]) + self.scale_range[0]
image, kps = cut_image(image_path, kps, scale, box[0], box[1])
ratio = 1.0 * args.crop_size / image.shape[0]
kps[:, :2] *= ratio
dst_image = cv2.resize(image, (args.crop_size, args.crop_size),
interpolation=cv2.INTER_CUBIC)
if self.use_flip and random.random() <= self.flip_prob:
dst_image, kps = flip_image(dst_image, kps)
#normalize kp to [-1, 1]
ratio = 1.0 / args.crop_size
kps[:, :2] = 2.0 * kps[:, :2] * ratio - 1.0
return {
'image':
torch.tensor(
convert_image_by_pixformat_normalize(dst_image,
self.pix_format,
self.normalize)).float(),
'kp_2d':
torch.tensor(kps).float(),
'image_name':
self.images[index],
'data_set':
'lsp'
}
def __getitem__(self, index):
imgPath = self.images[index]
keyps = self.keyps[index].copy()
bbox = self.boxes[index]
scale = np.random.rand(4) * (self.scaleRange[1] -
self.scaleRange[0]) + self.scaleRange[0]
bboxImg, bboxKps = util.cut_image(imgPath, keyps, scale, bbox[0],
bbox[1])
ratio = 1.0 * self.cropSize / bboxImg.shape[0]
bboxKps[:, :2] *= ratio
resImg = cv2.resize(bboxImg, (self.cropSize, self.cropSize),
interpolation=cv2.INTER_CUBIC)
ratio = 1.0 / self.cropSize
bboxKps[:, :2] = 2.0 * bboxKps[:, :2] * ratio - 1.0
return {
"image":
torch.tensor(
util.convert_image_by_pixformat_normalize(
resImg, self.pixFormat, self.normalize)).float(),
"kp_2d":
torch.tensor(bboxKps).float(),
"image_name":
imgPath,
"data_set":
"COCO 2017"
}
def __getitem__(self, index):
imgPath = self.images[index]
keyps2d = self.keyps2d[index].copy()
# remove head and neck points
keyps2d = keyps2d[0:12, :]
bbox = self.boxes[index]
keyps3d = self.keyps3d[index].copy()
keyps3d = keyps3d[0:12, :]
"""# plot points on image and save
imgVis=cv2.imread(imgPath)
for r in range(keyps2d.shape[0]):
currX=int(keyps2d[r,0])
currY=int(keyps2d[r,1])
cv2.circle(imgVis,(currX,currY),2,(0,0,255),4)
cv2.putText(imgVis,str(r),(currX,currY),cv2.FONT_HERSHEY_SIMPLEX,0.2,(255,255,255))
cv2.imwrite("/home/BaseCode/trainedModels/5/tmpFull.png",imgVis)"""
#scale=np.random.rand(4)*(self.scaleRange[1]-self.scaleRange[0])+self.scaleRange[0]
scale = 1
bboxImg, bboxKps = util.cut_image(imgPath, keyps2d, scale, bbox[0],
bbox[1])
bboxKps = bboxKps[:, 0:2]
bboxKps[:, 0] *= (1.0 * self.cropSize / bboxImg.shape[0])
bboxKps[:, 1] *= (1.0 * self.cropSize / bboxImg.shape[1])
resImg = cv2.resize(bboxImg, (self.cropSize, self.cropSize),
interpolation=cv2.INTER_CUBIC)
#bboxKps[:,:2]=2.0*bboxKps[:, :2]*ratio-1.0
bboxKps = (2.0 * bboxKps * 1.0 / self.cropSize) - 1.0
#print(bboxKps)
shape, pose = self.smplShapeParams[index], self.smplPoseParams[index]
#theta=np.concatenate((np.zeros(3),pose,shape),axis = 0)
theta = np.concatenate((pose, shape), axis=0)
return {
"image":
torch.from_numpy(
util.convert_image_by_pixformat_normalize(
resImg, self.pixFormat, self.normalize)).float(),
"kp_2d":
torch.from_numpy(bboxKps).float(),
"kp_3d":
torch.from_numpy(keyps3d).float(),
"theta":
torch.from_numpy(theta).float(),
"image_name":
self.images[index],
"w_smpl":
1.0,
"w_3d":
1.0,
"data_set":
"Human3.6M"
}
def __getitem__(self, index):
image_path = self.images[index]
kps = self.kp2ds[index].copy()
box = self.boxs[index]
kp_3d = self.kp3ds[index].copy()
scale = np.random.rand(4) * (self.scale_range[1] -
self.scale_range[0]) + self.scale_range[0]
image, kps = cut_image(image_path, kps, scale, box[0], box[1])
ratio = 1.0 * args.crop_size / image.shape[0]
kps[:, :2] *= ratio
dst_image = cv2.resize(image, (args.crop_size, args.crop_size),
interpolation=cv2.INTER_CUBIC)
trival, shape, pose = np.zeros(
3), self.shapes[index], self.poses[index]
if self.use_flip and random.random() <= self.flip_prob:
dst_image, kps = flip_image(dst_image, kps)
pose = reflect_pose(pose)
kp_3d = reflect_lsp_kp(kp_3d)
#normalize kp to [-1, 1]
ratio = 1.0 / args.crop_size
kps[:, :2] = 2.0 * kps[:, :2] * ratio - 1.0
theta = np.concatenate((trival, pose, shape), axis=0)
return {
'image':
torch.from_numpy(
convert_image_by_pixformat_normalize(dst_image,
self.pix_format,
self.normalize)).float(),
'kp_2d':
torch.from_numpy(kps).float(),
'kp_3d':
torch.from_numpy(kp_3d).float(),
'theta':
torch.from_numpy(theta).float(),
'image_name':
self.images[index],
'w_smpl':
1.0,
'w_3d':
1.0,
'data_set':
'hum3.6m'
}
