vkashko commited on
Commit
120a49b
1 Parent(s): ae2c7cf

refactor: script and readme

Browse files
Files changed (2) hide show
  1. README.md +70 -0
  2. portrait_and_26_photos.py +61 -26
README.md CHANGED
@@ -7,6 +7,76 @@ language:
7
  tags:
8
  - finance
9
  - code
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
10
  ---
11
 
12
  # The Portrait and 26 Photos (272 people)
 
7
  tags:
8
  - finance
9
  - code
10
+ dataset_info:
11
+ features:
12
+ - name: portrait_1
13
+ dtype: image
14
+ - name: photo_1
15
+ dtype: image
16
+ - name: photo_2
17
+ dtype: image
18
+ - name: photo_3
19
+ dtype: image
20
+ - name: photo_4
21
+ dtype: image
22
+ - name: photo_5
23
+ dtype: image
24
+ - name: photo_6
25
+ dtype: image
26
+ - name: photo_7
27
+ dtype: image
28
+ - name: photo_8
29
+ dtype: image
30
+ - name: photo_9
31
+ dtype: image
32
+ - name: photo_10
33
+ dtype: image
34
+ - name: photo_11
35
+ dtype: image
36
+ - name: photo_12
37
+ dtype: image
38
+ - name: photo_13
39
+ dtype: image
40
+ - name: photo_14
41
+ dtype: image
42
+ - name: photo_15
43
+ dtype: image
44
+ - name: photo_16
45
+ dtype: image
46
+ - name: photo_17
47
+ dtype: image
48
+ - name: photo_18
49
+ dtype: image
50
+ - name: photo_19
51
+ dtype: image
52
+ - name: photo_20
53
+ dtype: image
54
+ - name: photo_21
55
+ dtype: image
56
+ - name: photo_22
57
+ dtype: image
58
+ - name: photo_23
59
+ dtype: image
60
+ - name: photo_24
61
+ dtype: image
62
+ - name: photo_25
63
+ dtype: image
64
+ - name: photo_26
65
+ dtype: image
66
+ - name: worker_id
67
+ dtype: string
68
+ - name: age
69
+ dtype: int8
70
+ - name: country
71
+ dtype: string
72
+ - name: gender
73
+ dtype: string
74
+ splits:
75
+ - name: train
76
+ num_bytes: 927211725
77
+ num_examples: 14
78
+ download_size: 923699881
79
+ dataset_size: 927211725
80
  ---
81
 
82
  # The Portrait and 26 Photos (272 people)
portrait_and_26_photos.py CHANGED
@@ -3,7 +3,7 @@ import pandas as pd
3
 
4
  _CITATION = """\
5
  @InProceedings{huggingface:dataset,
6
- title = {FaceSegmentation},
7
  author = {TrainingDataPro},
8
  year = {2023}
9
  }
@@ -14,7 +14,7 @@ An example of a dataset that we've collected for a photo edit App.
14
  The dataset includes 20 selfies of people (man and women)
15
  in segmentation masks and their visualisations.
16
  """
17
- _NAME = 'FaceSegmentation'
18
 
19
  _HOMEPAGE = f"https://huggingface.co/datasets/TrainingDataPro/{_NAME}"
20
 
@@ -30,11 +30,37 @@ class FaceSegmentation(datasets.GeneratorBasedBuilder):
30
  return datasets.DatasetInfo(
31
  description=_DESCRIPTION,
32
  features=datasets.Features({
33
- 'image': datasets.Image(),
34
- 'mask': datasets.Image(),
35
- 'id': datasets.Value('string'),
36
- 'gender': datasets.Value('string'),
37
- 'age': datasets.Value('int8')
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
38
  }),
39
  supervised_keys=None,
40
  homepage=_HOMEPAGE,
@@ -43,34 +69,43 @@ class FaceSegmentation(datasets.GeneratorBasedBuilder):
43
 
44
  def _split_generators(self, dl_manager):
45
  images = dl_manager.download(f"{_DATA}images.tar.gz")
46
- masks = dl_manager.download(f"{_DATA}masks.tar.gz")
47
  annotations = dl_manager.download(f"{_DATA}{_NAME}.csv")
48
  images = dl_manager.iter_archive(images)
49
- masks = dl_manager.iter_archive(masks)
50
  return [
51
  datasets.SplitGenerator(name=datasets.Split.TRAIN,
52
  gen_kwargs={
53
  "images": images,
54
- 'masks': masks,
55
  'annotations': annotations
56
  }),
57
  ]
58
 
59
- def _generate_examples(self, images, masks, annotations):
60
- annotations_df = pd.read_csv(annotations, sep=';')
 
 
 
61
 
62
- for idx, ((image_path, image),
63
- (mask_path, mask)) in enumerate(zip(images, masks)):
64
- yield idx, {
65
- "image": {
66
- "path": image_path,
67
- "bytes": image.read()
68
- },
69
- "mask": {
70
- "path": mask_path,
71
- "bytes": mask.read()
72
- },
73
- 'id': annotations_df['id'].iloc[idx],
74
- 'gender': annotations_df['gender'].iloc[idx],
75
- 'age': annotations_df['age'].iloc[idx]
76
  }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
 
4
  _CITATION = """\
5
  @InProceedings{huggingface:dataset,
6
+ title = {portrait_and_26_photos},
7
  author = {TrainingDataPro},
8
  year = {2023}
9
  }
 
14
  The dataset includes 20 selfies of people (man and women)
15
  in segmentation masks and their visualisations.
16
  """
17
+ _NAME = 'portrait_and_26_photos'
18
 
19
  _HOMEPAGE = f"https://huggingface.co/datasets/TrainingDataPro/{_NAME}"
20
 
 
30
  return datasets.DatasetInfo(
31
  description=_DESCRIPTION,
32
  features=datasets.Features({
33
+ 'portrait_1': datasets.Image(),
34
+ 'photo_1': datasets.Image(),
35
+ 'photo_2': datasets.Image(),
36
+ 'photo_3': datasets.Image(),
37
+ 'photo_4': datasets.Image(),
38
+ 'photo_5': datasets.Image(),
39
+ 'photo_6': datasets.Image(),
40
+ 'photo_7': datasets.Image(),
41
+ 'photo_8': datasets.Image(),
42
+ 'photo_9': datasets.Image(),
43
+ 'photo_10': datasets.Image(),
44
+ 'photo_11': datasets.Image(),
45
+ 'photo_12': datasets.Image(),
46
+ 'photo_13': datasets.Image(),
47
+ 'photo_14': datasets.Image(),
48
+ 'photo_15': datasets.Image(),
49
+ 'photo_16': datasets.Image(),
50
+ 'photo_17': datasets.Image(),
51
+ 'photo_18': datasets.Image(),
52
+ 'photo_19': datasets.Image(),
53
+ 'photo_20': datasets.Image(),
54
+ 'photo_21': datasets.Image(),
55
+ 'photo_22': datasets.Image(),
56
+ 'photo_23': datasets.Image(),
57
+ 'photo_24': datasets.Image(),
58
+ 'photo_25': datasets.Image(),
59
+ 'photo_26': datasets.Image(),
60
+ 'worker_id': datasets.Value('string'),
61
+ 'age': datasets.Value('int8'),
62
+ 'country': datasets.Value('string'),
63
+ 'gender': datasets.Value('string')
64
  }),
65
  supervised_keys=None,
66
  homepage=_HOMEPAGE,
 
69
 
70
  def _split_generators(self, dl_manager):
71
  images = dl_manager.download(f"{_DATA}images.tar.gz")
 
72
  annotations = dl_manager.download(f"{_DATA}{_NAME}.csv")
73
  images = dl_manager.iter_archive(images)
 
74
  return [
75
  datasets.SplitGenerator(name=datasets.Split.TRAIN,
76
  gen_kwargs={
77
  "images": images,
 
78
  'annotations': annotations
79
  }),
80
  ]
81
 
82
+ def _generate_examples(self, images, annotations):
83
+ annotations_df = pd.read_csv(annotations, sep=',')
84
+ images_data = pd.DataFrame(columns=['Link', 'Bytes'])
85
+ for idx, (image_path, image) in enumerate(images):
86
+ images_data.loc[idx] = {'Link': image_path, 'Bytes': image.read()}
87
 
88
+ annotations_df = pd.merge(annotations_df, images_data)
89
+ for idx, worker_id in enumerate(pd.unique(annotations_df['WorkerId'])):
90
+ annotation = annotations_df.loc[annotations_df['WorkerId'] ==
91
+ worker_id]
92
+ annotation = annotation.sort_values(['Type'])
93
+ data = {
94
+ row[5]: {
95
+ 'path': row[6],
96
+ 'bytes': row[7]
97
+ } for row in annotation.itertuples()
 
 
 
 
98
  }
99
+
100
+ age = annotation.loc[annotation['Type'] ==
101
+ 'portrait_1']['Age'].values[0]
102
+ country = annotation.loc[annotation['Type'] ==
103
+ 'portrait_1']['Country'].values[0]
104
+ gender = annotation.loc[annotation['Type'] ==
105
+ 'portrait_1']['Gender'].values[0]
106
+
107
+ data['worker_id'] = worker_id
108
+ data['age'] = age
109
+ data['country'] = country
110
+ data['gender'] = gender
111
+ yield idx, data