def _load_images_salmaps(self, datapath=None, imgext="bmp"):
""" load preliminary data (images, segmentations, and salience maps) """
if datapath is None:
if sys.platform == "darwin":
homedir = "/Users/rzhao/"
else:
homedir = "/home/rzhao/"
datapath = homedir + "Dropbox/ongoing/reid_jrnl/salgt/data_viper/"
filepath = datapath + "query/"
imgfiles = sorted(glob(filepath + "*." + imgext))
imgs = [imread(im) for im in imgfiles]
salfilepath = datapath + "labels.pkl"
data = loadfile(salfilepath)
segmsks, salmsks = data
imgs = [imresize(im, size=(segmsks[0].shape), interp="bicubic") for im in imgs]
# imgs_norm = [imnormalize(im) for im in imgs]
# return imgs, segmsks, salmsks
self.imgs = imgs
self.segmsks = segmsks
self.salmsks = salmsks
def loadfile(file=None):
''' Load data from file with different format '''
if file is None:
raise NameError('File not specified!')
print 'Loading file at {}'.format(file)
if file[-3:] == '.gz':
f = gzip.open(file, 'rb')
data = cPickle.load(f)
elif file[-3:] == 'pkl':
with open(file, 'rb') as f:
data = cPickle.load(f)
elif file[-3:] == 'csv':
with open(file, 'rb') as f:
reader = csv.reader(f)
data = [row for row in reader]
elif file[-3:] == 'mat':
data = loadmat(file)
elif file[-3:] in ['jpg', 'png', 'bmp']:
data = imread(file)
else:
raise NameError('File format not recognized')
return data
def _load_images_salmaps(self, datapath=None, imgext="bmp"):
""" load preliminary data (images, segmentations, and salience maps) """
if datapath is None:
if sys.platform == "darwin":
homedir = "/Users/rzhao/"
else:
homedir = "/home/rzhao/"
datapath = homedir + "Dropbox/ongoing/reid_jrnl/salgt/data_viper/"
filepath = datapath + "query/"
imgfiles = sorted(glob(filepath + "*." + imgext))
self.nPerson = len(imgfiles)
imgs = [imread(im) for im in imgfiles]
salfilepath = datapath + "labels.pkl"
data = loadfile(salfilepath)
segmsks, salmsks = data
imgs = [imresize(im, size=(self.imH, self.imW)) for im in imgs]
segmsks = [imresize(im, size=(self.imH, self.imW)) for im in segmsks]
salmsks = [imresize(im, size=(self.imH, self.imW)) / 255.0 for im in salmsks]
self.imgs = np.asarray(imgs)
self.segmsks = np.asarray(segmsks)
self.salmsks = np.asarray(salmsks)
# load dense colorsift features
labeled_imidx_path = "../data_viper/labeled_imidx.mat"
tmp = loadfile(labeled_imidx_path)
labeled_imidx = tmp["labeled_imidx"].flatten()
feat_path = homedir + "Dropbox/ongoing/reid_jrnl/salgt/data_viper/features.mat"
tmp = loadfile(feat_path)
self.feats = tmp["features"].astype(np.float)[labeled_imidx]
def convert2pkl(self, pklfile):
if not os.path.isfile(pklfile):
dataset_dir = "/home/rzhao/Projects/deep-saliency/data/"
thus10000 = dataset_dir + "THUS10000_Imgs_GT/Imgs"
msra5000 = dataset_dir + "MSRA5000/Imgs"
msra5000_test = dataset_dir + "MSRA5000/MSRA-B-test"
img_ext = ".jpg"
msk_ext = ".png"
augX = 10
trn_img = []
trn_msk = []
for single_image in sorted(glob(thus10000 + "/*" + img_ext)):
rsb = glob(msra5000_test + "/*_" + ntpath.basename(single_image)[:-4] + "_smap" + msk_ext)
if len(rsb) == 0:
trn_img.append(single_image)
trn_msk.append(single_image[:-4] + msk_ext)
tst_img = []
tst_msk = []
for single_image in sorted(glob(msra5000_test + "/*" + msk_ext)):
tst_img.append(msra5000 + "/" + ntpath.basename(single_image)[: -len("_smap.png")] + img_ext)
tst_msk.append(msra5000 + "/" + ntpath.basename(single_image)[: -len("_smap.png")] + msk_ext)
# read images
print "reading ..."
train_img = [imread(fname) for fname in trn_img]
train_msk = [imread(fname) for fname in trn_msk]
test_img = [imread(fname) for fname in tst_img]
test_msk = [imread(fname) for fname in tst_msk]
# preprocessing
print "preprocessing ..."
train_x, train_y = self.preprocessing(train_img, train_msk, augx=augX)
test_x, test_y = self.preprocessing(test_img, test_msk, augx=0)
# shuffle training data
print "shuffle data ..."
np.random.seed(123)
np.random.shuffle(train_x)
np.random.seed(123)
np.random.shuffle(train_y)
# flattern and dtype conversion
print "flatten data ..."
train_x = np.asarray(train_x, dtype=np.float32)
train_y = np.asarray(train_y, dtype=np.float32)
test_x = np.asarray(test_x, dtype=np.float32)
test_y = np.asarray(test_y, dtype=np.float32)
train_x = imflatten(train_x)
train_y = imflatten(train_y)
test_x = imflatten(test_x)
test_y = imflatten(test_y)
# normalize data to have zero mean and unit std
train_x = normalize01(train_x)
test_x = normalize01(test_x)
# split into train and valid
nValid = np.int(len(train_img) * 0.1) * augX
train = [train_x[:-nValid], train_y[:-nValid]]
valid = [train_x[-nValid:], train_y[-nValid:]]
# train = [train_x[0:7000], train_y[0:7000]]
# valid = [train_x[7000:], train_y[7000:]]
test = [test_x, test_y]
data = [train, valid, test]
self.save(data, pklfile)
else:
print "History pickle file exists!"
def print_labeling(data_path = None):
# if data_path is None:
# newDialog = QDialog()
# fpath = QFileDialog.getExistingDirectory(newDialog, "Select data directory", '../')
# if len(fpath) == 0:
# QMessageBox.warning(None, 'Warning!', 'Nothing loaded.')
# return
# data_path = str(fpath) + '/' # loaded path
src_file = data_path + 'parts.pkl'
usr_file = sorted(glob(data_path + '#*.pkl'))
src = DataMan(src_file)
srcdata = src.load()
usrhits = []
for f in usr_file:
tmp = DataMan(f)
tmpdata = tmp.load()
usrhits.append(tmpdata['scores'])
save_path = data_path + 'result/'
qfiles = sorted(glob(data_path + 'query/*'))
im = imread(qfiles[0])
imsz = im.shape[0:2]
msk0 = np.zeros(srcdata['labels'][0].shape)
segmsks = []
salmsks = []
for i in range(len(qfiles)):
im = imread(qfiles[i])
seg = msk0.copy()
sal = msk0.copy()
for k in usrhits[0][i].keys():
idx = srcdata['labels'][i] == k
nhits = np.asarray([nhit[i][k] for nhit in usrhits])
sal[idx] = hit2score(nhits)
seg[idx] = k
salmsks.append(sal)
segmsks.append(seg)
# normalize all msk
# scaler = MinMaxScaler()
# salscores = scaler.fit_transform(np.asarray(salmsks))
# save label and salience score map
savefile([segmsks, salmsks], data_path + 'labels.pkl')
for i in range(len(qfiles)):
im = imread(qfiles[i])
msk = salmsks[i]*255.
im_rs = imresize(im, msk0.shape, interp='bicubic')
pl.figure(1)
pl.clf()
pl.subplot(1, 2, 1)
pl.imshow(im_rs)
pl.subplot(1, 2, 2)
pl.imshow(color.rgb2grey(im_rs), cmap='gray', alpha=0.6)
pl.imshow(msk, cmap='hot', vmin=0, vmax=255, alpha=0.6)
pl.savefig(save_path+'{0:03d}.jpg'.format(i))
print save_path +'{0:03d}.jpg'.format(i) + ' saved!'
visualize_imfolder(save_path)
