def request_chat():
request_data = request.get_json()
from_user = request_data.get('from_user', '')
to_user = request_data.get('to_user', '')
to_user_channel = "private-notification_user_%s" % (to_user)
from_user_channel = "private-notification_user_%s" % (from_user)
try:
bot = User.query.filter(User.id == to_user).one()
except NoResultFound:
print('Error! No bot (yet).')
except MultipleResultsFound:
print('Error! Wait what?')
try:
yolo.get_model(bot.username)
except:
print('Error! Cannot load model!')
# check if there is a channel that already exists between this two user
channel = Channel.query.filter( Channel.from_user.in_([from_user, to_user]) ) \
.filter( Channel.to_user.in_([from_user, to_user]) ) \
.first()
if not channel:
# Generate a channel...
chat_channel = "private-chat_%s_%s" % (from_user, to_user)
new_channel = Channel()
new_channel.from_user = from_user
new_channel.to_user = to_user
new_channel.name = chat_channel
db_session.add(new_channel)
db_session.commit()
else:
# Use the channel name stored on the database
chat_channel = channel.name
data = {
"from_user": from_user,
"to_user": to_user,
"from_user_notification_channel": from_user_channel,
"to_user_notification_channel": to_user_channel,
"channel_name": chat_channel,
}
# Trigger an event to the other user
pusher.trigger(to_user_channel, 'new_chat', data)
return jsonify(data)
def sample_graph():
# ======================================
# = graph =
# ======================================
if not os.path.exists(py.join(output_dir, args_.generator_pb)):
# model
Genc, Gdec, _ = module.get_model(args.model, n_atts, weight_decay=args.weight_decay)
# placeholders & inputs
xa = tf.placeholder(tf.float32, shape=[None, args.crop_size, args.crop_size, 3])
b_ = tf.placeholder(tf.float32, shape=[None, n_atts])
# sample graph
x = Gdec(Genc(xa, training=False), b_, training=False)
else:
# load freezed model
with tf.gfile.GFile(py.join(output_dir, args_.generator_pb), 'rb') as f:
graph_def = tf.GraphDef()
graph_def.ParseFromString(f.read())
tf.import_graph_def(graph_def, name='generator')
# placeholders & inputs
xa = sess.graph.get_tensor_by_name('generator/xa:0')
b_ = sess.graph.get_tensor_by_name('generator/b_:0')
# sample graph
x = sess.graph.get_tensor_by_name('generator/xb:0')
# ======================================
# = run function =
# ======================================
save_dir = '%s/output/%s/samples_testing_slide/%s_%s_%s_%s' % \
(args.flask_path, args.experiment_name, args.test_att_name, '{:g}'.format(args.test_int_min), '{:g}'.format(args.test_int_max), '{:g}'.format(args.test_int_step))
py.mkdir(save_dir)
def run():
cnt = 0
for _ in tqdm.trange(len_test_dataset):
# data for sampling
xa_ipt, a_ipt = sess.run(test_iter.get_next())
b_ipt = np.copy(a_ipt)
b__ipt = (b_ipt * 2 - 1).astype(np.float32) # !!!
x_opt_list = [xa_ipt]
for test_int in np.arange(args.test_int_min, args.test_int_max + 1e-5, args.test_int_step):
b__ipt[:, args.att_names.index(args.test_att_name)] = test_int
x_opt = sess.run(x, feed_dict={xa: xa_ipt, b_: b__ipt})
x_opt_list.append(x_opt)
sample = np.transpose(x_opt_list, (1, 2, 0, 3, 4))
sample = np.reshape(sample, (sample.shape[0], -1, sample.shape[2] * sample.shape[3], sample.shape[4]))
for s in sample:
cnt += 1
im.imwrite(s, '%s/%d.jpg' % (save_dir, cnt))
return run
def sample_graph():
# model
Genc, Gdec, _ = module.get_model(args.model, n_atts, weight_decay=args.weight_decay)
# placeholders & inputs
xa = tf.placeholder(tf.float32, shape=[None, args.crop_size, args.crop_size, 3], name='xa')
b_ = tf.placeholder(tf.float32, shape=[None, n_atts], name='b_')
# sample graph
x = Gdec(Genc(xa, training=False), b_, training=False)
x = tf.identity(x, name='xb')
def process_query(bot_id, file, name):
try:
bot = User.query.filter(User.id == bot_id).one()
except NoResultFound:
print('Error! No bot (yet).')
except MultipleResultsFound:
print('Error! Wait what?')
try:
model = yolo.get_model(bot.username)
except:
print('Error! Cannot get model!')
inp_dir = './data/input/{}.jpg'.format(name)
out_dir = './data/output/{}'.format(name)
file.save(inp_dir)
r = yolo.detect(model, inp_dir, out_dir=out_dir)
r = [x for x in r if x['prob'] > 0.3]
messages = []
messages.append('I have detected %d object(s) in the images.' % len(r))
for idx, det in enumerate(r):
prob = det['prob']
if prob > 0.8:
messages.append('[%d] I detected a [%s].' % (idx, det['class']))
elif prob > 0.5:
messages.append('[%d] I am not so sure but it may be a [%s]' %
(idx, det['class']))
elif prob > 0.3:
messages.append(
'[%d] I really do not know what this is. My best guess is that it is a [%s].'
% (idx, det['class']))
inp_token = upload_file('input', inp_dir)
out_token = upload_file('output', out_dir + '.jpg')
inp_url = 'https://firebasestorage.googleapis.com/v0/b/narupedia-d9cf9.appspot.com/o/input%2F' + name + '.jpg?alt=media&token=' + inp_token
out_url = 'https://firebasestorage.googleapis.com/v0/b/narupedia-d9cf9.appspot.com/o/output%2F' + name + '.jpg?alt=media&token=' + out_token
return inp_url, out_url, messages
def sample_graph():
# ======================================
# = graph =
# ======================================
test_next = test_iter.get_next()
if not os.path.exists(py.join(output_dir, 'generator.pb')):
# model
Genc, Gdec, _ = module.get_model(args.model,
n_atts,
weight_decay=args.weight_decay)
# placeholders & inputs
xa = tf.placeholder(tf.float32,
shape=[None, args.crop_size, args.crop_size, 3])
b_ = tf.placeholder(tf.float32, shape=[None, n_atts])
# sample graph
x = Gdec(Genc(xa, training=False), b_, training=False)
else:
# load freezed model
with tf.gfile.GFile(py.join(output_dir, 'generator.pb'), 'rb') as f:
graph_def = tf.GraphDef()
graph_def.ParseFromString(f.read())
tf.import_graph_def(graph_def, name='generator')
# placeholders & inputs
xa = sess.graph.get_tensor_by_name('generator/xa:0')
b_ = sess.graph.get_tensor_by_name('generator/b_:0')
# sample graph
x = sess.graph.get_tensor_by_name('generator/xb:0')
# ======================================
# = run function =
# ======================================
save_dir = './output/%s/samples_testing_multi' % args.experiment_name
tmp = ''
for test_att_name, test_int in zip(args.test_att_names, args.test_ints):
tmp += '_%s_%s' % (test_att_name, '{:g}'.format(test_int))
save_dir = py.join(save_dir, tmp[1:])
py.mkdir(save_dir)
def run():
cnt = 0
for _ in tqdm.trange(len_test_dataset):
# data for sampling
xa_ipt, a_ipt = sess.run(test_next)
b_ipt = np.copy(a_ipt)
for test_att_name in args.test_att_names:
i = args.att_names.index(test_att_name)
b_ipt[..., i] = 1 - b_ipt[..., i]
b_ipt = data.check_attribute_conflict(b_ipt, test_att_name,
args.att_names)
b__ipt = (b_ipt * 2 - 1).astype(np.float32) # !!!
for test_att_name, test_int in zip(args.test_att_names,
args.test_ints):
i = args.att_names.index(test_att_name)
b__ipt[..., i] = b__ipt[..., i] * test_int
x_opt_list = [xa_ipt]
x_opt = sess.run(x, feed_dict={xa: xa_ipt, b_: b__ipt})
x_opt_list.append(x_opt)
sample = np.transpose(x_opt_list, (1, 2, 0, 3, 4))
sample = np.reshape(sample, (sample.shape[0], -1, sample.shape[2] *
sample.shape[3], sample.shape[4]))
for s in sample:
cnt += 1
im.imwrite(s, '%s/%d.jpg' % (save_dir, cnt))
return run
def sample_graph():
# ======================================
# = graph =
# ======================================
if not os.path.exists(py.join(output_dir, 'generator.pb')):
# model
Genc, Gdec, _ = module.get_model(args.model,
n_atts,
weight_decay=args.weight_decay)
# placeholders & inputs
xa = tf.placeholder(tf.float32,
shape=[None, args.crop_size, args.crop_size, 3])
b_ = tf.placeholder(tf.float32, shape=[None, n_atts])
# sample graph
x = Gdec(Genc(xa, training=False), b_, training=False)
else:
# load freezed model
with tf.gfile.GFile(py.join(output_dir, 'generator.pb'), 'rb') as f:
graph_def = tf.GraphDef()
graph_def.ParseFromString(f.read())
tf.import_graph_def(graph_def, name='generator')
# placeholders & inputs
xa = sess.graph.get_tensor_by_name('generator/xa:0')
b_ = sess.graph.get_tensor_by_name('generator/b_:0')
# sample graph
x = sess.graph.get_tensor_by_name('generator/xb:0')
# ======================================
# = run function =
# ======================================
save_dir = '%s/output/%s/samples_testing_%s' % (
args.flask_path, args.experiment_name, '{:g}'.format(args.test_int))
py.mkdir(save_dir)
def run():
cnt = 0
for _ in tqdm.trange(len_test_dataset):
# data for sampling
xa_ipt, a_ipt = sess.run(test_iter.get_next())
b_ipt_list = [a_ipt] # the first is for reconstruction
for i in range(n_atts):
tmp = np.array(a_ipt, copy=True)
tmp[:, i] = 1 - tmp[:, i] # inverse attribute
tmp = data.check_attribute_conflict(tmp, args.att_names[i],
args.att_names)
b_ipt_list.append(tmp)
x_opt_list = [xa_ipt]
for i, b_ipt in enumerate(b_ipt_list):
b__ipt = (b_ipt * 2 - 1).astype(np.float32) # !!!
if i > 0: # i == 0 is for reconstruction
b__ipt[..., i - 1] = b__ipt[..., i - 1] * args.test_int
x_opt = sess.run(x, feed_dict={xa: xa_ipt, b_: b__ipt})
x_opt_list.append(x_opt)
sample = np.transpose(x_opt_list, (1, 2, 0, 3, 4))
sample = np.reshape(sample, (sample.shape[0], -1, sample.shape[2] *
sample.shape[3], sample.shape[4]))
for s in sample:
cnt += 1
im.imwrite(s, '%s/%d.jpg' % (save_dir, cnt))
return run
# ==============================================================================
# = data and model =
# ==============================================================================
# data
train_dataset, len_train_dataset = data.make_celeba_dataset(args.img_dir, args.train_label_path, args.att_names, args.batch_size,
load_size=args.load_size, crop_size=args.crop_size,
training=True, shuffle=True, repeat=None)
val_dataset, len_val_dataset = data.make_celeba_dataset(args.img_dir, args.val_label_path, args.att_names, args.n_samples,
load_size=args.load_size, crop_size=args.crop_size,
training=False, shuffle=True, repeat=None)
train_iter = train_dataset.make_one_shot_iterator()
val_iter = val_dataset.make_one_shot_iterator()
# model
Genc, Gdec, D = module.get_model(args.model, n_atts, weight_decay=args.weight_decay)
# loss functions
d_loss_fn, g_loss_fn = tfprob.get_adversarial_losses_fn(args.adversarial_loss_mode)
# ==============================================================================
# = graph =
# ==============================================================================
def D_train_graph():
# ======================================
# = graph =
# ======================================
# placeholders & inputs
def send_file():
request_data = request.form
from_user = request_data.get('from_user', '')
to_user = request_data.get('to_user', '')
channel = request_data.get('channel')
file = request.files['file']
bot_id = to_user
name = channel + datetime.datetime.now().strftime("%Y%m%d%H%M%S")
inp_dir = './data/input/{}.jpg'.format(name)
file.save(inp_dir)
inp_token = upload_file('input', inp_dir)
inp_url = 'https://firebasestorage.googleapis.com/v0/b/narupedia-d9cf9.appspot.com/o/input%2F' + name + '.jpg?alt=media&token=' + inp_token
inp_img = create_message(message=inp_url,
type='Image',
to_user=to_user,
from_user=from_user,
channel=channel)
pusher.trigger(channel, 'new_message', inp_img)
message = create_message(message="I'm onto it, please wait for a bit.",
type='Text',
to_user=from_user,
from_user=to_user,
channel=channel)
pusher.trigger(channel, 'new_message', message)
try:
bot = User.query.filter(User.id == bot_id).one()
except NoResultFound:
print('Error! No bot (yet).')
except MultipleResultsFound:
print('Error! Wait what?')
try:
model = yolo.get_model(bot.username)
except:
print('Error! Cannot get model!')
out_dir = './data/output/{}'.format(name)
r = yolo.detect(model, inp_dir, out_dir=out_dir)
r = [x for x in r if x['prob'] > 0.5]
messages = []
messages.append('I have detected %d object(s) in the images.' % len(r))
for idx, det in enumerate(r):
prob = det['prob']
if prob > 0.8:
messages.append('[%d] I detected a [%s].' % (idx, det['class']))
elif prob > 0.5:
messages.append('[%d] I am not so sure but it may be a [%s]' %
(idx, det['class']))
# elif prob > 0.3:
# messages.append('[%d] I really do not know what this is. My best guess is that it is a [%s].' % (idx, det['class']))
out_token = upload_file('output', out_dir + '.jpg')
out_url = 'https://firebasestorage.googleapis.com/v0/b/narupedia-d9cf9.appspot.com/o/output%2F' + name + '.jpg?alt=media&token=' + out_token
# inp_url, out_url, messages = process_query(bot_id, file, name)
out_img = create_message(message=out_url,
type='Image',
to_user=from_user,
from_user=to_user,
channel=channel)
pusher.trigger(channel, 'new_message', out_img)
for m in messages:
message = create_message(message=m,
type='Text',
to_user=from_user,
from_user=to_user,
channel=channel)
pusher.trigger(channel, 'new_message', message)
return jsonify(messages)
