Add at new repo again
This commit is contained in:
Binary file not shown.
Binary file not shown.
201
vton-api/preprocess/humanparsing/datasets/datasets.py
Normal file
201
vton-api/preprocess/humanparsing/datasets/datasets.py
Normal file
@@ -0,0 +1,201 @@
|
||||
#!/usr/bin/env python
|
||||
# -*- encoding: utf-8 -*-
|
||||
|
||||
"""
|
||||
@Author : Peike Li
|
||||
@Contact : peike.li@yahoo.com
|
||||
@File : datasets.py
|
||||
@Time : 8/4/19 3:35 PM
|
||||
@Desc :
|
||||
@License : This source code is licensed under the license found in the
|
||||
LICENSE file in the root directory of this source tree.
|
||||
"""
|
||||
|
||||
import os
|
||||
import numpy as np
|
||||
import random
|
||||
import torch
|
||||
import cv2
|
||||
from torch.utils import data
|
||||
from utils.transforms import get_affine_transform
|
||||
|
||||
|
||||
class LIPDataSet(data.Dataset):
|
||||
def __init__(self, root, dataset, crop_size=[473, 473], scale_factor=0.25,
|
||||
rotation_factor=30, ignore_label=255, transform=None):
|
||||
self.root = root
|
||||
self.aspect_ratio = crop_size[1] * 1.0 / crop_size[0]
|
||||
self.crop_size = np.asarray(crop_size)
|
||||
self.ignore_label = ignore_label
|
||||
self.scale_factor = scale_factor
|
||||
self.rotation_factor = rotation_factor
|
||||
self.flip_prob = 0.5
|
||||
self.transform = transform
|
||||
self.dataset = dataset
|
||||
|
||||
list_path = os.path.join(self.root, self.dataset + '_id.txt')
|
||||
train_list = [i_id.strip() for i_id in open(list_path)]
|
||||
|
||||
self.train_list = train_list
|
||||
self.number_samples = len(self.train_list)
|
||||
|
||||
def __len__(self):
|
||||
return self.number_samples
|
||||
|
||||
def _box2cs(self, box):
|
||||
x, y, w, h = box[:4]
|
||||
return self._xywh2cs(x, y, w, h)
|
||||
|
||||
def _xywh2cs(self, x, y, w, h):
|
||||
center = np.zeros((2), dtype=np.float32)
|
||||
center[0] = x + w * 0.5
|
||||
center[1] = y + h * 0.5
|
||||
if w > self.aspect_ratio * h:
|
||||
h = w * 1.0 / self.aspect_ratio
|
||||
elif w < self.aspect_ratio * h:
|
||||
w = h * self.aspect_ratio
|
||||
scale = np.array([w * 1.0, h * 1.0], dtype=np.float32)
|
||||
return center, scale
|
||||
|
||||
def __getitem__(self, index):
|
||||
train_item = self.train_list[index]
|
||||
|
||||
im_path = os.path.join(self.root, self.dataset + '_images', train_item + '.jpg')
|
||||
parsing_anno_path = os.path.join(self.root, self.dataset + '_segmentations', train_item + '.png')
|
||||
|
||||
im = cv2.imread(im_path, cv2.IMREAD_COLOR)
|
||||
h, w, _ = im.shape
|
||||
parsing_anno = np.zeros((h, w), dtype=np.long)
|
||||
|
||||
# Get person center and scale
|
||||
person_center, s = self._box2cs([0, 0, w - 1, h - 1])
|
||||
r = 0
|
||||
|
||||
if self.dataset != 'test':
|
||||
# Get pose annotation
|
||||
parsing_anno = cv2.imread(parsing_anno_path, cv2.IMREAD_GRAYSCALE)
|
||||
if self.dataset == 'train' or self.dataset == 'trainval':
|
||||
sf = self.scale_factor
|
||||
rf = self.rotation_factor
|
||||
s = s * np.clip(np.random.randn() * sf + 1, 1 - sf, 1 + sf)
|
||||
r = np.clip(np.random.randn() * rf, -rf * 2, rf * 2) if random.random() <= 0.6 else 0
|
||||
|
||||
if random.random() <= self.flip_prob:
|
||||
im = im[:, ::-1, :]
|
||||
parsing_anno = parsing_anno[:, ::-1]
|
||||
person_center[0] = im.shape[1] - person_center[0] - 1
|
||||
right_idx = [15, 17, 19]
|
||||
left_idx = [14, 16, 18]
|
||||
for i in range(0, 3):
|
||||
right_pos = np.where(parsing_anno == right_idx[i])
|
||||
left_pos = np.where(parsing_anno == left_idx[i])
|
||||
parsing_anno[right_pos[0], right_pos[1]] = left_idx[i]
|
||||
parsing_anno[left_pos[0], left_pos[1]] = right_idx[i]
|
||||
|
||||
trans = get_affine_transform(person_center, s, r, self.crop_size)
|
||||
input = cv2.warpAffine(
|
||||
im,
|
||||
trans,
|
||||
(int(self.crop_size[1]), int(self.crop_size[0])),
|
||||
flags=cv2.INTER_LINEAR,
|
||||
borderMode=cv2.BORDER_CONSTANT,
|
||||
borderValue=(0, 0, 0))
|
||||
|
||||
if self.transform:
|
||||
input = self.transform(input)
|
||||
|
||||
meta = {
|
||||
'name': train_item,
|
||||
'center': person_center,
|
||||
'height': h,
|
||||
'width': w,
|
||||
'scale': s,
|
||||
'rotation': r
|
||||
}
|
||||
|
||||
if self.dataset == 'val' or self.dataset == 'test':
|
||||
return input, meta
|
||||
else:
|
||||
label_parsing = cv2.warpAffine(
|
||||
parsing_anno,
|
||||
trans,
|
||||
(int(self.crop_size[1]), int(self.crop_size[0])),
|
||||
flags=cv2.INTER_NEAREST,
|
||||
borderMode=cv2.BORDER_CONSTANT,
|
||||
borderValue=(255))
|
||||
|
||||
label_parsing = torch.from_numpy(label_parsing)
|
||||
|
||||
return input, label_parsing, meta
|
||||
|
||||
|
||||
class LIPDataValSet(data.Dataset):
|
||||
def __init__(self, root, dataset='val', crop_size=[473, 473], transform=None, flip=False):
|
||||
self.root = root
|
||||
self.crop_size = crop_size
|
||||
self.transform = transform
|
||||
self.flip = flip
|
||||
self.dataset = dataset
|
||||
self.root = root
|
||||
self.aspect_ratio = crop_size[1] * 1.0 / crop_size[0]
|
||||
self.crop_size = np.asarray(crop_size)
|
||||
|
||||
list_path = os.path.join(self.root, self.dataset + '_id.txt')
|
||||
val_list = [i_id.strip() for i_id in open(list_path)]
|
||||
|
||||
self.val_list = val_list
|
||||
self.number_samples = len(self.val_list)
|
||||
|
||||
def __len__(self):
|
||||
return len(self.val_list)
|
||||
|
||||
def _box2cs(self, box):
|
||||
x, y, w, h = box[:4]
|
||||
return self._xywh2cs(x, y, w, h)
|
||||
|
||||
def _xywh2cs(self, x, y, w, h):
|
||||
center = np.zeros((2), dtype=np.float32)
|
||||
center[0] = x + w * 0.5
|
||||
center[1] = y + h * 0.5
|
||||
if w > self.aspect_ratio * h:
|
||||
h = w * 1.0 / self.aspect_ratio
|
||||
elif w < self.aspect_ratio * h:
|
||||
w = h * self.aspect_ratio
|
||||
scale = np.array([w * 1.0, h * 1.0], dtype=np.float32)
|
||||
|
||||
return center, scale
|
||||
|
||||
def __getitem__(self, index):
|
||||
val_item = self.val_list[index]
|
||||
# Load training image
|
||||
im_path = os.path.join(self.root, self.dataset + '_images', val_item + '.jpg')
|
||||
im = cv2.imread(im_path, cv2.IMREAD_COLOR)
|
||||
h, w, _ = im.shape
|
||||
# Get person center and scale
|
||||
person_center, s = self._box2cs([0, 0, w - 1, h - 1])
|
||||
r = 0
|
||||
trans = get_affine_transform(person_center, s, r, self.crop_size)
|
||||
input = cv2.warpAffine(
|
||||
im,
|
||||
trans,
|
||||
(int(self.crop_size[1]), int(self.crop_size[0])),
|
||||
flags=cv2.INTER_LINEAR,
|
||||
borderMode=cv2.BORDER_CONSTANT,
|
||||
borderValue=(0, 0, 0))
|
||||
input = self.transform(input)
|
||||
flip_input = input.flip(dims=[-1])
|
||||
if self.flip:
|
||||
batch_input_im = torch.stack([input, flip_input])
|
||||
else:
|
||||
batch_input_im = input
|
||||
|
||||
meta = {
|
||||
'name': val_item,
|
||||
'center': person_center,
|
||||
'height': h,
|
||||
'width': w,
|
||||
'scale': s,
|
||||
'rotation': r
|
||||
}
|
||||
|
||||
return batch_input_im, meta
|
||||
@@ -0,0 +1,89 @@
|
||||
#!/usr/bin/env python
|
||||
# -*- encoding: utf-8 -*-
|
||||
|
||||
"""
|
||||
@Author : Peike Li
|
||||
@Contact : peike.li@yahoo.com
|
||||
@File : dataset.py
|
||||
@Time : 8/30/19 9:12 PM
|
||||
@Desc : Dataset Definition
|
||||
@License : This source code is licensed under the license found in the
|
||||
LICENSE file in the root directory of this source tree.
|
||||
"""
|
||||
|
||||
import os
|
||||
import pdb
|
||||
|
||||
import cv2
|
||||
import numpy as np
|
||||
from PIL import Image
|
||||
from torch.utils import data
|
||||
from utils.transforms import get_affine_transform
|
||||
|
||||
|
||||
class SimpleFolderDataset(data.Dataset):
|
||||
def __init__(self, root, input_size=[512, 512], transform=None):
|
||||
self.root = root
|
||||
self.input_size = input_size
|
||||
self.transform = transform
|
||||
self.aspect_ratio = input_size[1] * 1.0 / input_size[0]
|
||||
self.input_size = np.asarray(input_size)
|
||||
self.is_pil_image = False
|
||||
if isinstance(root, Image.Image):
|
||||
self.file_list = [root]
|
||||
self.is_pil_image = True
|
||||
elif os.path.isfile(root):
|
||||
self.file_list = [os.path.basename(root)]
|
||||
self.root = os.path.dirname(root)
|
||||
else:
|
||||
self.file_list = os.listdir(self.root)
|
||||
|
||||
def __len__(self):
|
||||
return len(self.file_list)
|
||||
|
||||
def _box2cs(self, box):
|
||||
x, y, w, h = box[:4]
|
||||
return self._xywh2cs(x, y, w, h)
|
||||
|
||||
def _xywh2cs(self, x, y, w, h):
|
||||
center = np.zeros((2), dtype=np.float32)
|
||||
center[0] = x + w * 0.5
|
||||
center[1] = y + h * 0.5
|
||||
if w > self.aspect_ratio * h:
|
||||
h = w * 1.0 / self.aspect_ratio
|
||||
elif w < self.aspect_ratio * h:
|
||||
w = h * self.aspect_ratio
|
||||
scale = np.array([w, h], dtype=np.float32)
|
||||
return center, scale
|
||||
|
||||
def __getitem__(self, index):
|
||||
if self.is_pil_image:
|
||||
img = np.asarray(self.file_list[index])[:, :, [2, 1, 0]]
|
||||
else:
|
||||
img_name = self.file_list[index]
|
||||
img_path = os.path.join(self.root, img_name)
|
||||
img = cv2.imread(img_path, cv2.IMREAD_COLOR)
|
||||
h, w, _ = img.shape
|
||||
|
||||
# Get person center and scale
|
||||
person_center, s = self._box2cs([0, 0, w - 1, h - 1])
|
||||
r = 0
|
||||
trans = get_affine_transform(person_center, s, r, self.input_size)
|
||||
input = cv2.warpAffine(
|
||||
img,
|
||||
trans,
|
||||
(int(self.input_size[1]), int(self.input_size[0])),
|
||||
flags=cv2.INTER_LINEAR,
|
||||
borderMode=cv2.BORDER_CONSTANT,
|
||||
borderValue=(0, 0, 0))
|
||||
|
||||
input = self.transform(input)
|
||||
meta = {
|
||||
'center': person_center,
|
||||
'height': h,
|
||||
'width': w,
|
||||
'scale': s,
|
||||
'rotation': r
|
||||
}
|
||||
|
||||
return input, meta
|
||||
@@ -0,0 +1,40 @@
|
||||
import torch
|
||||
from torch.nn import functional as F
|
||||
|
||||
|
||||
def generate_edge_tensor(label, edge_width=3):
|
||||
label = label.type(torch.cuda.FloatTensor)
|
||||
if len(label.shape) == 2:
|
||||
label = label.unsqueeze(0)
|
||||
n, h, w = label.shape
|
||||
edge = torch.zeros(label.shape, dtype=torch.float).cuda()
|
||||
# right
|
||||
edge_right = edge[:, 1:h, :]
|
||||
edge_right[(label[:, 1:h, :] != label[:, :h - 1, :]) & (label[:, 1:h, :] != 255)
|
||||
& (label[:, :h - 1, :] != 255)] = 1
|
||||
|
||||
# up
|
||||
edge_up = edge[:, :, :w - 1]
|
||||
edge_up[(label[:, :, :w - 1] != label[:, :, 1:w])
|
||||
& (label[:, :, :w - 1] != 255)
|
||||
& (label[:, :, 1:w] != 255)] = 1
|
||||
|
||||
# upright
|
||||
edge_upright = edge[:, :h - 1, :w - 1]
|
||||
edge_upright[(label[:, :h - 1, :w - 1] != label[:, 1:h, 1:w])
|
||||
& (label[:, :h - 1, :w - 1] != 255)
|
||||
& (label[:, 1:h, 1:w] != 255)] = 1
|
||||
|
||||
# bottomright
|
||||
edge_bottomright = edge[:, :h - 1, 1:w]
|
||||
edge_bottomright[(label[:, :h - 1, 1:w] != label[:, 1:h, :w - 1])
|
||||
& (label[:, :h - 1, 1:w] != 255)
|
||||
& (label[:, 1:h, :w - 1] != 255)] = 1
|
||||
|
||||
kernel = torch.ones((1, 1, edge_width, edge_width), dtype=torch.float).cuda()
|
||||
with torch.no_grad():
|
||||
edge = edge.unsqueeze(1)
|
||||
edge = F.conv2d(edge, kernel, stride=1, padding=1)
|
||||
edge[edge!=0] = 1
|
||||
edge = edge.squeeze()
|
||||
return edge
|
||||
Reference in New Issue
Block a user