def bag_of_words(self, words, vocab):
words = b.rem_special_chars(words)
words = b.convert_to_lowcase(words)
words = b.rem_stopwords(words)
words = b.lemma(words)
words = self.text_to_word(words)
bag = np.zeros(len(vocab))
for w in words:
for i, x in enumerate(vocab):
if x == w:
bag[i] += 1
return np.array(bag)
def read_part_sift_descriptors(rate=0.05):
"""
read_part_sift_descriptors select descriptors from a certain percent, indicated by rate, of
sift features of images from each class. Actually, the descriptors are corresponding to
that of read_part_sift_features_images(rate)
For example,
Codes:
part_sift_descriptors = read_part_sift_descriptors()
print(part_sift_descriptors.shape)
Outputs:
(3236336, 128)
"""
import BOW
return BOW.read_list_sift_features(rate)
def process_df_frames_train_set(item, method='surf', histogram_norm='l1'):
import FeatureExtraction
import random
filename, df_frames = item
if df_frames['frame_index'].iloc[0] % 100 != 24:
return []
boxes_dict = Preprocessing.extract_box_img(df_frames)
train_set = []
kmeans = BOW.load_kmeans(f'kmeans_{method}_128.pkl')
for category, boxes in boxes_dict.items():
for box in boxes:
_, des = FeatureExtraction.feature_extraction(box, method)
if des is not None:
histogram = BOW.compute_histogram(des,
kmeans,
method=method,
norm=histogram_norm)
train_set.append((histogram, category))
return train_set
def __init__(self):
self.bowHOG = BOW()
self.bowHOF = BOW()
self.bowMBFx = BOW()
self.bowMBFy = BOW()
self.dimHOG = 96
self.dimHOF = 108
self.dimMBFx = 96
self.dimMBFy = 96
self.vocszHOG = 128
self.vocszHOF = 128
self.vocszMBFx = 128
self.vocszMBFy = 128
class DenseTrajBOW:
def __init__(self):
self.bowHOG = BOW()
self.bowHOF = BOW()
self.bowMBFx = BOW()
self.bowMBFy = BOW()
self.dimHOG = 96
self.dimHOF = 108
self.dimMBFx = 96
self.dimMBFy = 96
self.vocszHOG = 128
self.vocszHOF = 128
self.vocszMBFx = 128
self.vocszMBFy = 128
def build(self,dataHOG,dataHOF,dataMBFx,dataMBFy):
self.bowHOG.vq(data=dataHOG,voc_size=self.vocszHOG,gt_labels=None)
def calcFeatures(self,dataHOG,dataHOF,dataMBFx,dataMBFy):
self.bowHOG.calc_bow_representation(fv=dataHOG)
return self.bowHOG.bow
def read_bowed_labeled_features(use_dl=False, use_cv=False):
"""
use_dl: use deep learning features or BOW features
:return a list [features, labels]
"""
if use_dl:
features = read_features(use_dl=True)
labels = []
for line in open(global_defs.PATH_LABELS, 'r'):
labels.append(int(line))
labels = np.array(labels)
else:
if not use_cv:
import BOW
return BOW.read_BOWed_labeled_features()
else:
import sift_bow_cv
return sift_bow_cv.read_labeled_BOWed_features_cv()
return [features, labels]
words = b.convert_to_lowcase(words)
words = b.rem_stopwords(words)
words = b.lemma(words)
words = self.text_to_word(words)
bag = np.zeros(len(vocab))
for w in words:
for i, x in enumerate(vocab):
if x == w:
bag[i] += 1
return np.array(bag)
# Read data from a file
f = open('text2.txt', 'r')
corpus = f.read()
corpus = b.rem_special_chars(corpus)
corpus = b.convert_to_lowcase(corpus)
corpus = b.rem_stopwords(corpus)
corpus = b.lemma(corpus)
t = Tokenizer()
s = t.text_to_word(corpus)
print("word count=", t.word_count(s))
print("number of unique words", t.number_of_unique_words(s))
inp = "autonomous individuals mutual aid self governance"
print("input = ", inp)
print("bag of words = ", t.bag_of_words(inp, s))
def clfs_rolling_windows(img,
clf,
stride=8,
padding=0,
ratios=((1, 1), (3, 1), (1, 3)),
scale=1.5,
prob_threshold=0.75,
min_side_length=16,
method='sift',
kmeans_path=None):
if method not in ('sift', 'surf'):
raise ValueError("method param should be sift or surf")
if kmeans_path is None:
kmeans_path = f'kmeans_{method}_128.pkl'
h, w = img.shape[:2]
padded_image = np.zeros((h + padding * 2, w + padding * 2, img.shape[2]),
dtype='uint8')
padded_image[padding:h + padding, padding:w + padding, :] = img
h, w = padded_image.shape[:2]
box_hist_dict = {
'box': [],
'histogram': [],
'high_prob': [],
'high_category': []
}
kmeans = BOW.load_kmeans(kmeans_path)
kp, des = FeatureExtraction.feature_extraction(img, method)
df_kp = pd.DataFrame({
'x': [item.pt[0] for item in kp],
'y': [item.pt[1] for item in kp]
})
for ratio in ratios:
print(f'ratio {ratio}')
ratio_h, ratio_w = ratio
if ratio_h < ratio_w:
# (h, w)
start_window = (min_side_length,
int(min_side_length * ratio_w / ratio_h))
else:
start_window = (int(min_side_length * ratio_h / ratio_w),
min_side_length)
window = start_window
this_stride = stride
tmp = window
count = 0
while tmp[0] <= h and tmp[1] <= w:
count += 1
tmp = (int(tmp[0] * scale), int(tmp[1] * scale))
pbar = tqdm(total=count)
while window[0] <= h and window[1] <= w:
for i in range(0, h - window[0], this_stride):
for j in range(0, w - window[1], this_stride):
#print(i, j)
box = (i, j, i + window[0], j + window[1])
#img_this = padded_image[box[0]:box[2], box[1]:box[3]]
#_, des_this = FeatureExtraction.feature_extraction(img_this, method)
des_this = des[np.logical_and.reduce(
(box[0] < df_kp['y'], df_kp['y'] < box[2],
box[1] < df_kp['x'], df_kp['x'] < box[3]))]
if des_this is None or len(des_this) == 0:
continue
histogram = BOW.compute_histogram(des_this,
kmeans,
method=method,
norm='l0')
if histogram is None:
continue
#print(clf.predict(histogram.reshape(1, -1)))
"""cv2.namedWindow('image', cv2.WINDOW_AUTOSIZE)
cv2.imshow('image', img_this)
cv2.waitKey()
cv2.destroyAllWindows()"""
box_hist_dict['box'].append(box)
box_hist_dict['histogram'].append(histogram)
window = (int(window[0] * scale), int(window[1] * scale))
this_stride = int(this_stride * np.sqrt(scale))
pbar.update(1)
pbar.close()
X = np.array(box_hist_dict['histogram'])
y_prob = clf.predict_proba(X)
high_prob = np.max(y_prob, axis=1).ravel().tolist()
high_category = np.argmax(y_prob, axis=1).ravel().tolist()
box_hist_dict['high_prob'] = high_prob
box_hist_dict['high_category'] = high_category
df_box_hist = pd.DataFrame(box_hist_dict)
df_box_hist = df_box_hist[df_box_hist['high_category'] != 0]
df_box_hist = df_box_hist[df_box_hist['high_prob'] > prob_threshold]
return df_box_hist
#B_gt_labels = np.ones(N_B,dtype='int') #feat = np.concatenate((A,B), axis=0) #gt_labels = np.concatenate((A_gt_labels,B_gt_labels), axis=0) # Test Case 2: Tiny dataset # desc = []; desc.append([1,1]) desc.append([1.5,1]) desc.append([1,15]) desc.append([1.5,1]) desc.append([1,1.5]) desc.append([1,1.6]) desc.append([10,10]) desc.append([12,10]) desc.append([10,13]) desc.append([14,10]) desc.append([10,15]) feat = sp.vstack(tuple(desc)) gt_labels = np.array([0,0,1,0,0,0,1,1,1,1,1]) V = BOW(data=feat,voc_size=2,gt_labels=gt_labels) #print V.vq_data feat2 = np.concatenate((feat,feat),axis=0) V.calc_bow_representation(fv=feat2) print V.bow
