def predict():
if request.method == "POST":
response_object = validate_token()
if response_object['status'] == 'success':
try:
video = request.files['video']
plates_like_objects, image_binary, car_image = check_video(
video, app.config['UPLOAD_FOLDER'])
result_dict = get_placa(plates_like_objects, image_binary,
model)
prediction_id = uuid.uuid4()
result_img = Image.fromarray(car_image).convert('RGB')
# Guardar la imagen
result_img.save("./result_img/" + str(prediction_id) + ".jpg")
# TODO guardar predicción , ruta de imagen en BD y usuario en tabla de BD,para listarlo luego en Historial
nwPrediccion = historial(user_id=response_object['email'],
placa=result_dict['placa'],
foto=str(prediction_id))
db.session.add(nwPrediccion)
db.session.commit()
base64img = encode(car_image)
# limpiar carpeta output y uploads
delete_files_in_directory(app.config['UPLOAD_FOLDER'])
delete_files_in_directory('output')
result_dict["img"] = base64img
return make_response(jsonify(result_dict)), 200
except Exception:
return make_response(jsonify({"status": "failure"})), 400
else:
return make_response(jsonify(response_object)), 401
def get_content():
vh_title = "yohoho."
list_li = []
VALUES["kinopoisk"] = _kp_id_
response = tools.get_response(URL, HEADERS, VALUES, 'POST')
if response:
jdata = json.loads(response)
for host in ENABLED_HOSTS:
host_data = jdata[host]
if host_data:
iframe = host_data["iframe"]
translate = host_data["translate"]
quality = host_data["quality"]
#{"vodlocker":{},
# "hdgo":{"iframe":"https://hdgo.cx/video/oSlSCtQ0t8apv6vJGD1va2xbKTd9k8YC/17223/","translate":"Дублированный","quality":"плохое TS"},
# "iframe":{"iframe":"https://videoframe.at/movie/2eb6408pc8p/iframe","translate":"Полное дублирование","quality":"TS"},
# "torrent":{"iframe":"https://4h0y.yohoho.cc/?title=%D1%85%D0%B8%D1%89%D0%BD%D0%B8%D0%BA"},
# "hdbaza":{"iframe":"https://vidozzz.com/iframe?mh=bbd8ed61c2256ea4&uh=65bd8ef1126daa6f","translate":"Viruseproject","quality":""},
# "kodik":{"iframe":"https://kodik.cc/video/15298/6f7fcc06b4e7d51f4ff574af5a59115e/720p","translate":"Проф. Многоголосый","quality":"BDRip 720p"},
# "trailer":{"iframe":"https://hdgo.cx/video/trailer/oSlSCtQ0t8apv6vJGD1va2xbKTd9k8YC/17223/"},
# "moonwalk":{"iframe":"https://streamguard.cc/video/d9419273b3fea0ef15980f70e35cc078/iframe?show_translations=1","translate":"Дубляж","quality":""}}
title_ = "*T*"
title = "[COLOR=orange][{0}][/COLOR] {1} ({2})".format(
vh_title + host, tools.encode(title_),
translate + "," + quality)
uri = sys.argv[0] + "?mode=show&url={0}".format(
urllib.quote_plus(prepare_url(host, iframe)))
item = xbmcgui.ListItem(title)
list_li.append([uri, item, True])
return list_li
def get_content(part):
vh_title = "hdgo.club"
list_li = []
VALUES["search"] = _kp_id_
HEADERS["Referer"] = URL + part
response = get_response(URL + part, HEADERS, VALUES, 'POST')
if response:
try:
arr = common.parseDOM(response, "div", attrs={"class": "li_gen"})
for item in arr:
try:
divs = common.parseDOM(item, "div")
url_ = "http:" + common.parseDOM(item, "a", attrs={"class": "btn-primary"}, ret="href")[0]
except:
continue
url = prepare_url(url_)
title_ = tools.strip(divs[1].split("\n")[0]) + add_title_info(divs)
title = "[COLOR=orange][{0}][/COLOR] {1}".format(vh_title, tools.encode(title_))
uri = sys.argv[0] + "?mode=show&url={0}&title={1}".format(urllib.quote_plus(url), urllib.quote_plus(title))
item = xbmcgui.ListItem(title)
list_li.append([uri, item, True])
except:
pass
return list_li
def callALPR(image):
base64image = encode(image)
SECRET_KEY = 'sk_d6c76f5e23014eff289f1313'
url = 'https://api.openalpr.com/v3/recognize_bytes?recognize_vehicle=1&country=us&secret_key=%s' % (
SECRET_KEY)
r = requests.post(url, data=base64image, verify=False)
output_dict = r.json()
return {"placa": output_dict['results'][0]['plate'], "flag_alpr": True}
def user_history():
response_object = validate_token()
if response_object['status'] == 'success':
lst_predicciones = historial.query.filter_by(user_id="xyz")\
.order_by(historial.id.desc()).all() #response_object["email"])
for prediccion in lst_predicciones:
car_image_output = imread("./result_img/" + prediccion.foto +
".jpg")
base64img = encode(car_image_output)
prediccion.foto = base64img
return make_response(
jsonify(results=[elem.serialize()
for elem in lst_predicciones])), 200
return make_response(jsonify(response_object))
def evaluate(model, seed=1234, evaltest=True):
"""
Run experiment
"""
print 'Preparing data...'
train, dev, test, scores = load_data()
train[0], train[1], scores[0] = shuffle(train[0],
train[1],
scores[0],
random_state=seed)
print 'Computing training skipthoughts...'
trainA = tools.encode(model, train[0], verbose=False, use_eos=True)
trainB = tools.encode(model, train[1], verbose=False, use_eos=True)
print 'Computing development skipthoughts...'
devA = tools.encode(model, dev[0], verbose=False, use_eos=True)
devB = tools.encode(model, dev[1], verbose=False, use_eos=True)
print 'Computing feature combinations...'
trainF = np.c_[np.abs(trainA - trainB), trainA * trainB]
devF = np.c_[np.abs(devA - devB), devA * devB]
print 'Encoding labels...'
trainY = encode_labels(scores[0])
devY = encode_labels(scores[1])
with tf.device('/cpu:0'):
print 'Compiling model...'
lrmodel = prepare_model(ninputs=trainF.shape[1])
print 'Training...'
bestlrmodel = train_model(lrmodel, trainF, trainY, devF, devY,
scores[1])
if evaltest:
print 'Computing test skipthoughts...'
testA = tools.encode(model, test[0], verbose=False, use_eos=True)
testB = tools.encode(model, test[1], verbose=False, use_eos=True)
print 'Computing feature combinations...'
testF = np.c_[np.abs(testA - testB), testA * testB]
print 'Evaluating...'
r = np.arange(1, 6)
yhat = np.dot(bestlrmodel.predict_proba(testF, verbose=2), r)
pr = pearsonr(yhat, scores[2])[0]
sr = spearmanr(yhat, scores[2])[0]
se = mse(yhat, scores[2])
print 'Test Pearson: ' + str(pr)
print 'Test Spearman: ' + str(sr)
print 'Test MSE: ' + str(se)
return yhat, pr, sr, se
def get_content():
vh_title = "czx.to"
list_li = []
response = tools.get_response(URL + '/' + str(_kp_id_) + '/', HEADERS,
VALUES, 'GET')
if response:
iframe = common.parseDOM(response, "iframe", ret="src")[0]
title_ = "*T*"
title = "[COLOR=orange][{0}][/COLOR] {1}".format(
vh_title, tools.encode(title_))
uri = sys.argv[0] + "?mode=show&url={0}".format(
urllib.quote_plus(iframe))
item = xbmcgui.ListItem(title)
list_li.append([uri, item, True])
return list_li
def summarize_trained(self, sentences):
# Getting sentence embeddings
vectors = tools.encode(self.model, sentences, verbose=False)
print('Sentences have been encoded...')
# Retrieving clusters
n_clusters = int(np.ceil(len(vectors)**0.5))
# n_clusters = int(np.ceil(SUMMARY_LENGTH))
kmeans = KMeans(n_clusters=n_clusters, random_state=0)
# print pca embeddings
self.print_embeddings(vectors)
kmeans.fit(vectors)
avg = []
for j in range(n_clusters):
idx = np.where(kmeans.labels_ == j)[0]
avg.append(np.mean(idx))
# Choosing sentences closest to cluster centers
closest, _ = pairwise_distances_argmin_min(kmeans.cluster_centers_,
vectors)
ordering = sorted(range(n_clusters), key=lambda k: avg[k])
# Returning summary
summary = ' '.join([sentences[closest[idx]] for idx in ordering])
return summary
def session_start(self):
self.session = vk_api.VkApi(login=self.login,
password=self.password,
auth_handler=two_factor_auth_handler,
captcha_handler=captcha_handler,
config_filename='vk_config.json')
clear_last_log()
self.set_logfile_name()
Logging("Trying to start session", self.logfile)
self.session.auth()
Logging("Session started", self.logfile)
Logging("Getting music api", self.logfile)
self.api = self.session.get_api()
self.music = audio.VkAudio(self.session)
Logging("Complete successfully", self.logfile)
self.get_id()
with open('user.json', 'w') as f:
user_data = {
'login': self.login,
'password': tools.encode(self.password)
}
json.dump(user_data, f, indent=2, ensure_ascii=False)
return (self)
def get_content(part):
vh_title = "kodik.top"
list_li = []
VALUES["search"] = _kp_id_
HEADERS["Referer"] = URL + part
response = tools.get_response(URL + part, HEADERS, VALUES, 'POST')
if response:
try:
table = common.parseDOM(response,
"table",
attrs={"class": "table table-hover"})
tbody = common.parseDOM(table, "tbody")
rows = common.parseDOM(tbody, "tr")
for item in rows:
try:
tds = common.parseDOM(item, "td")
url_ = "https:" + common.parseDOM(
item,
"a",
attrs={"class": "btn btn-success btn-xs copypreview"},
ret="data-link")[0]
except:
continue
url = prepare_url(url_)
title_ = tools.strip(tds[0]) + " (" + tools.strip(
tds[1]) + ", " + tools.strip(tds[2]) + ")"
title = "[COLOR=orange][{0}][/COLOR] {1}".format(
vh_title, tools.encode(title_))
uri = sys.argv[0] + "?mode=show&url={0}&title={1}".format(
urllib.quote_plus(url), urllib.quote_plus(title))
item = xbmcgui.ListItem(title)
list_li.append([uri, item, True])
except:
pass
return list_li
# Author: Shalin Luo
# Apr 14, 2016
# File to loading model from tools.py and encode new sentence
import tools
input_fpath = 'data/ch2v_list'
output_fpath = 'data/vector_output'
embed_map = tools.load_googlenews_vectors()
model = tools.load_model(embed_map)
# X = ['l o v e ', 'r u n ', 'l i v e', 'e n d']
word_list = []
with open(input_fpath) as f:
word = f.readline().strip()
while word != '':
word_list.append(word)
word = f.readline().strip()
vector_list = tools.encode(model, word_list)
# output
with open(output_fpath, 'a') as f_output:
for vector in vector_list:
output_str = (" ".join(str(elem) for elem in vector)) + '\n'
f_output.write(output_str)
print '*****END*****'
def run(self):
thumbnail_dpath = os.path.join(self.clip_dpath, 'media')
try:
tools.mkdir_p(thumbnail_dpath)
except OSError as err:
wx.CallAfter(
self.wx_parent.error_message,
_('Permission error creating clip directory.')
)
return
tools.save_safe_image(self.clip_dpath, self.bg_input_fpath)
tools.save_safe_image(self.clip_dpath, self.pip_input_fpath)
tools.save_thumbnails(thumbnail_dpath, self.pip_input_fpath)
pip_width, pip_height = tools.get_pip_size(
self.pip_input_fpath,
self.pip_scale
)
bg_fname = tools.sanitize_path(os.path.split(self.bg_input_fpath)[1])
pip_fname = tools.sanitize_path(os.path.split(self.pip_input_fpath)[1])
clip_config = tools.default_clip_config()
clip_config['name'] = self.clip_name
clip_config['data'] = bg_fname
clip_config['pip'] = pip_fname
clip_config['pipwidth'] = pip_width
clip_config['pipheight'] = pip_height
clip_config['category'] = self.clip_category
clip_config_buf = ''
# name, data, pip must be first 3 items in order
ordered_config_items = ['name', 'data', 'pip']
for line in ordered_config_items:
clip_config_buf += line + ' = ' + clip_config[line] + os.linesep
# Add the rest of the items whose order probably doesn't matter
for key, value in clip_config.iteritems():
key = unicode(key)
value = unicode(value)
if value and (key not in ordered_config_items):
clip_config_buf += key + ' = ' + value + os.linesep
clip_config_fpath = os.path.join(self.clip_dpath, 'playlist.sos')
with open(clip_config_fpath, 'wb') as clip_config_fobj:
clip_config_fobj.write(tools.encode(clip_config_buf))
# # Update the SOS library
# # Not used because updating the library doesn't automatically update
# # the playlist panel in SOS Stream GUI, and updating the playlist
# # panel in SOS Stream GUI the right way would be tedious.
# # Rescanning the library automatically without refreshing the
# # playlist panel could confuse users.
# child = subprocess.Popen(
# ['/shared/sos/default/bin/scan_library'], stdout=subprocess.PIPE
# )
# while child.poll() is None: # Read command output as it comes
# out = child.stdout.readline()
# print(out.strip()) # Put this in a status window later
# # The menu can be updated, at least:
# wish = Tkinter.Tk()
# wish.withdraw() # Hide Tk window. Possible with constructor?
# try:
# wish.send('sos_stream_gui',
# 'destroy',
# '.menubar.library')
# wish.send('sos_stream_gui',
# 'mk_library_menu',
# '.menubar.library')
# except Tkinter.TclError as err:
# pass # SOS Stream GUI not open
# Done
success_message = '\n'.join(
[
_('Clip "{}" successfully added.'),
_('To activate the changes, '
'perform the following actions in SOS Stream GUI:'),
_('Library -> Update Library...\n'
'Library -> (all/{})')
]
).format(
clip_config['name'],
clip_config['category']
)
wx.CallAfter(self.wx_parent.information_message, success_message)
with open(f) as file_descriptor:
file_content = file_descriptor.read().decode("utf-8", "ignore")
file_content = sent_tokenize(file_content)
for sentence in file_content:
if sentence:
X.append(sentence.strip())
sentence_embeddings = np.empty([file_counter, 4800])
loc = base_path_to_directory + "dictionary.pkl"
saveto = base_path_to_directory + "toy.npz"
maxlen_w = 70
worddict, wordcount = vocab.build_dictionary(X)
vocab.save_dictionary(worddict, wordcount,
loc) #loc where you want to save dictionary
#in train.py set 1>path for dictionary, 2>save_to -path where to save model 3>maxlen_w
train.trainer(X, dictionary=loc, saveto=saveto, maxlen_w=maxlen_w)
#In tools.py set path_to_model=save_to in train, path_to_dictionary=dictionary in train and path_to_word2vec.
embed_map = tools.load_googlenews_vectors(path_to_word2vec)
model = tools.load_model(embed_map)
if not os.path.exists(SENTENCE_EMBEDDING_FOLDER):
os.mkdir(SENTENCE_EMBEDDING_FOLDER)
for f in FILES:
with open(f) as file_descriptor:
file_content = sent_tokenize(file_descriptor.read())
document_embedding = tools.encode(model, file_content, verbose=False)
document_embedding = np.average(document_embedding, axis=0)
file_name = f.split('/')[-1]
np.save(SENTENCE_EMBEDDING_FOLDER + file_name[:-4], document_embedding)
def turn(self) -> (str, int, int):
# Fill these fields to return
message: str = None
message_value: int = 0
direction: int = Direction.CENTER.value
ant = self.game.ant
# print(ant.health)
x = ant.currentX
y = ant.currentY
base_x = self.game.baseX
base_y = self.game.baseY
self.turn_number = self.turn_number + 1
self.init_dirs()
if not self.vision:
# print("Creating vision")
for i in range(self.game.mapWidth):
new_line = []
for j in range(self.game.mapHeight):
new_line.append([(UNKNOWN, -1)])
self.vision.append(new_line)
# self.vision[self.game.mapWidth-base_x-1][self.game.mapHeight-base_y-1].append((ENEMY_BASE, self.turn_number))
cur = self.game.ant.visibleMap.cells[ant.currentX][ant.currentY]
for chat in self.game.chatBox.allChats:
msg = chat.text
for i in range(0, len(msg), 2):
cx, cy, obj = encode(msg[i:i+2])
if not has_obj(obj, self.vision[cx][cy]):
self.vision[cx][cy].append((obj, chat.turn))
new_objs = []
def upd(i, j, obj):
if not has_obj(obj, self.vision[i][j]):
new_objs.append((i, j, obj))
if cur.resource_type == ResourceType.BREAD.value or cur.resource_type == ResourceType.GRASS.value:
tools.last_resource = (x,y)
if cur.type == CellType.TRAP.value and tools.has_resource == 1:
# self.vision[tools.last_resource[0]][tools.last_resource[1]].append((WALL,self.turn_number))
upd(x, y, BAD_TRAP)
self.vision[x][y].append((BAD_TRAP,self.turn_number))
tools.allied_in_range = 0
tools.enemy_in_range = 0
tools.has_resource = 1 if ant.currentResource and ant.currentResource.value > 0 else 0
# print(self.turn_number , " " , tools.has_resource)
if self.turn_number == 0:
random_moves = []
for i in range(self.game.mapWidth):
for j in range(self.game.mapHeight):
cell = self.game.ant.visibleMap.cells[i][j]
if not cell:
continue
if cell.type == CellType.WALL.value:
upd(i, j, WALL)
self.vision[i][j].append((WALL, self.turn_number))
elif cell.type == CellType.TRAP.value:
upd(i, j, TRAP)
if self.turn_number == 0:
random_moves.append(cell)
self.vision[i][j].append((TRAP, self.turn_number))
elif cell.type == CellType.SWAMP.value:
upd(i, j, SWAMP)
self.vision[i][j].append((SWAMP, self.turn_number))
elif cell.type != CellType.EMPTY and cell.type == CellType.BASE.value and (base_x != i or base_y != j): # what
self.vision[i][j].append((ENEMY_BASE, self.turn_number))
else:
if self.turn_number == 0:
random_moves.append(cell)
if cell.resource_type == ResourceType.BREAD.value:
self.vision[i][j].append((BREAD, self.turn_number))
elif cell.resource_type == ResourceType.GRASS.value:
self.vision[i][j].append((GRASS, self.turn_number))
else:
self.vision[i][j].append((EMPTY, self.turn_number))
if cell.ants:
maximum = TEAM_KARGAR
for a in cell.ants:
if a.antType == AntType.KARGAR.value and a.antTeam == AntTeam.ALLIED.value:
this = TEAM_KARGAR
tools.allied_in_range += 0.5
elif a.antType == AntType.SARBAAZ.value and a.antTeam == AntTeam.ALLIED.value:
this = TEAM_SARBAZ
tools.allied_in_range += 1
elif a.antType == AntType.KARGAR.value and a.antTeam == AntTeam.ENEMY.value:
this = ENEMY_KARGAR
tools.enemy_in_range += 0.5
else:
this = ENEMY_SARBAZ
tools.enemy_in_range += 1
if this > maximum:
maximum = this
self.vision[i][j].append((maximum, self.turn_number))
else:
self.vision[i][j].append((NO_ANTS, self.turn_number))
self.vision[i][j] = prune(self.vision[i][j])
if self.turn_number == 0:
print(random_moves)
self.first_target = random.choice(random_moves)
AI.first_target = self.first_target
# print("turn: ", self.turn_number)
set_turn_number(self.turn_number)
if ant.antType == AntType.SARBAAZ.value:
if self.turn_number == 0 and len(cur.ants) > 1:
direction = random.choice([UP,DOWN,LEFT,RIGHT,CENTER])
else:
direction = self.get_move('scorpion')
elif ant.antType == AntType.KARGAR.value:
# if self.turn_number == 0 or self.turn_number == 1:
# direction = random.choice([UP,DOWN,LEFT,RIGHT,CENTER])
# else:
direction = self.get_move('ant')
AI.vision = self.vision
AI.turn_number = self.turn_number
message = ""
message_value = 1
for i, j, obj in new_objs:
message += decode(i, j, obj)
if len(message) > MAX_CHARS:
message = message[:MAX_CHARS]
return message, message_value, direction
embed_map = tools.load_googlenews_vectors()
model = tools.load_model(embed_map)
with open(model_save_path, 'w') as f:
pkl.dump(model, f)
else:
model = pkl.load(open(model_save_path, 'r'))
# process
xp = []
for entity in ent:
#line = re.sub(r'[0-9]+','#'," ".join([word for word in entity.split('_') if word not in stops])).lower()
line = " ".join([word for word in entity.split('_') if word not in stops]).lower()
xp.append(filter(lambda x: x in string.printable, line))
# encode new sentences
vectors = tools.encode(model, xp)
with open(vec_save_path, 'w') as f:
np.save(f, vectors)
else:
vectors = np.load(open(vec_save_path,'r'))
#print len(vectors)
#print len(labels)
#print ent[0]
#print labels[0]
#print set(labels)
# labels -> int
le = LabelEncoder()
lab = le.fit_transform(labels)
def encode(encoder, sentences, verbose=False):
sys.path.insert(0, 'training/')
import tools
return tools.encode(encoder, sentences)
def get_bin(self):
return encode(self.rank + self.suit)
# coding: utf-8
import vocab
import train
import tools
import numpy as np
with open("../../wikipedia_txt/result_wakati.txt") as f:
fdata = [line.rstrip() for i, line in enumerate(f)]
print '# lines: ', len(fdata)
worddict, wordcount = vocab.build_dictionary(fdata)
vocab.save_dictionary(worddict, wordcount, "word_dict")
print '# vocab: ', len(worddict)
train.trainer(fdata, dictionary="word_dict", saveFreq=100, saveto="model", reload_=True, n_words=40000)
model = tools.load_model()
vectors = tools.encode(model, fdata, use_norm=False)
np.savez('vecs.npz', vectors)
