-
Notifications
You must be signed in to change notification settings - Fork 0
/
data.py
58 lines (47 loc) · 2.48 KB
/
data.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
import os
import torch
import torch.nn as nn
import random
import torchvision.transforms as transforms
from torch.utils.data import DataLoader, Dataset
from PIL import Image
from parameters import *
import matplotlib.pyplot as plt
class TripletFaceDataset(Dataset):
def __init__(self, root_dir, transform = None):
super(TripletFaceDataset, self).__init__()
self.root_dir = root_dir
self.transform = transform
self.person_folders = [folder for folder in os.listdir(root_dir) if os.path.isdir(os.path.join(root_dir, folder))]
def __len__(self) -> int:
return len(self.person_folders)
def _get_images_from_folder(self, folderPath):
images = [img_name for img_name in os.listdir(folderPath) if img_name.endswith('.jpg')]
return images
def _random_triplet_indices(self, numImages):
anchor_idx, positive_idx = torch.randperm(numImages)[:2]
return anchor_idx, positive_idx
def __getitem__(self, index):
person_folder = self.person_folders[index]
person_path = os.path.join(self.root_dir, person_folder)
images = self._get_images_from_folder(person_path)
num_images = len(images)
if num_images < 2:
dummy_image = Image.new('RGB', (96, 96))
if self.transform:
dummy_image = self.transform(dummy_image)
return dummy_image, dummy_image, dummy_image
anchor_idx, positive_idx = self._random_triplet_indices(num_images)
anchor_path = os.path.join(person_path, images[anchor_idx])
positive_path = os.path.join(person_path, images[positive_idx])
negative_people = [self.person_folders[i] for i in range(len(self.person_folders)) if i != index and len(self._get_images_from_folder(os.path.join(self.root_dir, self.person_folders[i]))) >= 1]
negative_person = random.choice(negative_people)
negative_path = os.path.join(self.root_dir, negative_person, random.choice(os.listdir(os.path.join(self.root_dir, negative_person))))
anchor_image = Image.open(anchor_path).convert("RGB")
positive_image = Image.open(positive_path).convert("RGB")
negative_image = Image.open(negative_path).convert("RGB")
if self.transform:
anchor_image = self.transform(anchor_image)
positive_image = self.transform(positive_image)
negative_image = self.transform(negative_image)
return anchor_image, positive_image, negative_image