-
Notifications
You must be signed in to change notification settings - Fork 0
/
data_load.py
143 lines (113 loc) · 4.98 KB
/
data_load.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
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
import json
import os
import string
import torch
import torch.utils.data as data
from PIL import Image
from torchvision import datasets
from cocoapi2.PythonAPI.pycocotools.coco import COCO
class CocoDataset(data.Dataset):
"""COCO Custom Dataset compatible with torch.utils.data.DataLoader."""
def __init__(self, root, json, vocab, transform=None):
"""Set the path for images, captions and vocabulary wrapper.
Args:
root: image directory.
json: coco annotation file path.
vocab: vocabulary wrapper.
transform: image transformer.
"""
self.root = root
self.coco = COCO(json)
self.ids = list(self.coco.anns.keys())
self.vocab = vocab
self.transform = transform
def __getitem__(self, index):
"""Returns one data pair ( image, caption, image_id )."""
coco = self.coco
vocab = self.vocab
ann_id = self.ids[index]
caption = coco.anns[ann_id]['caption']
img_id = coco.anns[ann_id]['image_id']
filename = coco.loadImgs(img_id)[0]['file_name']
path = 'train2014/' + filename
image = Image.open(os.path.join(self.root, path)).convert('RGB')
if self.transform is not None:
image = self.transform(image)
# Convert caption (string) to word ids.
tokens = str(caption).lower().translate(string.punctuation).strip().split()
caption = []
caption.append(vocab('<start>'))
caption.extend([vocab(token) for token in tokens])
caption.append(vocab('<end>'))
target = torch.Tensor(caption)
return image, target, img_id, filename
def __len__(self):
return len(self.ids)
def collate_fn(data):
"""Creates mini-batch tensors from the list of tuples (image, caption).
We should build custom collate_fn rather than using default collate_fn,
because merging caption (including padding) is not supported in default.
Args:
data: list of tuple (image, caption).
- image: torch tensor of shape (3, 256, 256).
- caption: torch tensor of shape (?); variable length.
Returns:
images: torch tensor of shape (batch_size, 3, 256, 256).
targets: torch tensor of shape (batch_size, padded_length).
lengths: list; valid length for each padded caption.
img_ids: image ids in COCO dataset, for evaluation purpose
filenames: image filenames in COCO dataset, for evaluation purpose
"""
# Sort a data list by caption length (descending order).
data.sort(key=lambda x: len(x[1]), reverse=True)
images, captions, img_ids, filenames = zip(*data) # unzip
# Merge images (from tuple of 3D tensor to 4D tensor).
images = torch.stack(images, 0)
img_ids = list(img_ids)
filenames = list(filenames)
# Merge captions (from tuple of 1D tensor to 2D tensor).
lengths = [len(cap) for cap in captions]
targets = torch.zeros(len(captions), max(lengths)).long()
for i, cap in enumerate(captions):
end = lengths[i]
targets[i, :end] = cap[:end]
return images, targets, lengths, img_ids, filenames
def get_loader(root, json, vocab, transform, batch_size, shuffle, num_workers):
"""Returns torch.utils.data.DataLoader for custom coco dataset."""
# COCO caption dataset
coco = CocoDataset(root=root,
json=json,
vocab=vocab,
transform=transform)
# Data loader for COCO dataset
# This will return (images, captions, lengths) for every iteration.
# images: tensor of shape (batch_size, 3, 224, 224).
# captions: tensor of shape (batch_size, padded_length).
# lengths: list indicating valid length for each caption. length is (batch_size).
data_loader = torch.utils.data.DataLoader(dataset=coco,
batch_size=batch_size,
shuffle=shuffle,
num_workers=num_workers,
collate_fn=collate_fn)
return data_loader
class CocoEvalLoader(datasets.ImageFolder):
def __init__(self, root, ann_path, transform=None,
loader=datasets.folder.default_loader):
'''
Customized COCO loader to get Image ids and Image Filenames
root: path for images
ann_path: path for the annotation file (e.g., caption_val2014.json)
'''
self.folder = root
self.transform = transform
self.loader = loader
self.samples = json.load(open(ann_path, 'r'))['images']
def __getitem__(self, index):
filename = self.samples[index]['file_name']
img_id = self.samples[index]['id']
# Filename for the image
path = os.path.join(self.folder, 'val2014', filename)
img = Image.open(path).convert('RGB')
if self.transform is not None:
img = self.transform(img)
return img, img_id, filename