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
|
"""Modified Olivetti faces dataset.
The original database was available from (now defunct)
http://www.uk.research.att.com/facedatabase.html
The version retrieved here comes in MATLAB format from the personal
web page of Sam Roweis:
http://www.cs.nyu.edu/~roweis/
There are ten different images of each of 40 distinct subjects. For some
subjects, the images were taken at different times, varying the lighting,
facial expressions (open / closed eyes, smiling / not smiling) and facial
details (glasses / no glasses). All the images were taken against a dark
homogeneous background with the subjects in an upright, frontal position (with
tolerance for some side movement).
The original dataset consisted of 92 x 112, while the Roweis version
consists of 64x64 images.
"""
# Copyright (c) 2011 David Warde-Farley <wardefar at iro dot umontreal dot ca>
# License: Simplified BSD
from os.path import join, exists
from os import makedirs
from cStringIO import StringIO
import urllib2
import numpy as np
from scipy.io.matlab import loadmat
from .base import get_data_home, Bunch
from ..utils import check_random_state
from ..externals import joblib
DATA_URL = "http://cs.nyu.edu/~roweis/data/olivettifaces.mat"
TARGET_FILENAME = "olivetti.pkz"
# Grab the module-level docstring to use as a description of the
# dataset
MODULE_DOCS = __doc__
def fetch_olivetti_faces(data_home=None, shuffle=False, random_state=0,
download_if_missing=True):
"""Loader for the Olivetti faces data-set from AT&T.
Parameters
----------
data_home : optional, default: None
Specify another download and cache folder for the datasets. By default
all scikit learn data is stored in '~/scikit_learn_data' subfolders.
shuffle : boolean, optional
If True the order of the dataset is shuffled to avoid having
images of the same person grouped.
download_if_missing: optional, True by default
If False, raise a IOError if the data is not locally available
instead of trying to download the data from the source site.
random_state : optional, integer or RandomState object
The seed or the random number generator used to shuffle the
data.
Notes
------
This dataset consists of 10 pictures each of 40 individuals. The original
database was available from (now defunct)
http://www.uk.research.att.com/facedatabase.html
The version retrieved here comes in MATLAB format from the personal
web page of Sam Roweis:
http://www.cs.nyu.edu/~roweis/
"""
data_home = get_data_home(data_home=data_home)
if not exists(data_home):
makedirs(data_home)
if not exists(join(data_home, TARGET_FILENAME)):
print 'downloading Olivetti faces from %s to %s' % (DATA_URL,
data_home)
fhandle = urllib2.urlopen(DATA_URL)
buf = StringIO(fhandle.read())
mfile = loadmat(buf)
faces = mfile['faces'].T.copy()
joblib.dump(faces, join(data_home, TARGET_FILENAME), compress=6)
del mfile
else:
faces = joblib.load(join(data_home, TARGET_FILENAME))
# We want floating point data, but float32 is enough (there is only
# one byte of precision in the original uint8s anyway)
faces = np.float32(faces)
faces = faces - faces.min()
faces /= faces.max()
faces = faces.reshape((400, 64, 64)).transpose(0, 2, 1)
# 10 images per class, 400 images total, each class is contiguous.
target = np.array([i // 10 for i in range(400)])
if shuffle:
random_state = check_random_state(random_state)
order = random_state.permutation(len(faces))
faces = faces[order]
target = target[order]
return Bunch(data=faces.reshape(len(faces), -1),
images=faces,
target=target,
DESCR=MODULE_DOCS)
|
