def main(_argv):
model = ArcFaceModel(size=FLAGS.input_size, training=False, use_pretrain=False)
model.load_weights(FLAGS.weights).expect_partial()
model.summary()
# Saved path will be 'output_dir/model_name/version'
saved_path = os.path.join(FLAGS.output_dir, 'arcface', str(FLAGS.version))
tf.saved_model.save(model, saved_path)
logging.info("model saved to: {}".format(saved_path))
model = tf.saved_model.load(saved_path)
infer = model.signatures[tf.saved_model.DEFAULT_SERVING_SIGNATURE_DEF_KEY]
logging.info(infer.structured_outputs)
if not FLAGS.image:
return
img = tf.image.decode_image(open(FLAGS.image, 'rb').read(), channels=3)
img = tf.image.resize(img, (FLAGS.input_size, FLAGS.input_size))
img = img / 255.
img = tf.expand_dims(img, 0)
t1 = time.time()
outputs = infer(img)
embeddings = outputs['OutputLayer']
t2 = time.time()
logging.info('time: {}'.format(t2 - t1))
def main():
cfg = load_yaml('./configs/arc_res50_mask.yaml')
model = ArcFaceModel(size=cfg['input_size'],
backbone_type=cfg['backbone_type'],
num_classes=cfg['num_classes'],
head_type=cfg['head_type'],
embd_shape=cfg['embd_shape'],
w_decay=cfg['w_decay'],
training=False)
model.summary()
ckpt_path = tf.train.latest_checkpoint('./checkpoints/' + cfg['sub_name'])
if ckpt_path is not None:
print("[*] load ckpt from {}".format(ckpt_path))
model.load_weights(ckpt_path)
else:
print("[*] training from scratch.")
temp1 = np.ones((62,112,3))
temp2 = np.zeros((50,112,3))
masked_img = np.concatenate([temp1, temp2], axis =0)
path_img1 = '/home/anhdq23/Desktop/nguyen/data/AR/test2/M-002-12.bmp'
path_img2 = '/home/anhdq23/Desktop/nguyen/data/AR/test2/M-003-01.bmp'
img1 = Image.open(path_img1)
img1 = img1.resize((112, 112))
img1 = np.array(img1)/255.0
img2 = Image.open(path_img2)
img2 = img2.resize((112, 112))
img2 = np.array(img2)/255.0
mask_img2 = np.multiply(img2, masked_img)
fc1 = model.predict(mask_img2.reshape((1,112,112,3)))
norm_fc1 = preprocessing.normalize(fc1.reshape((1,512)), norm='l2', axis=1)
fc2 = model.predict(img2.reshape((1,112,112,3)))
norm_fc2 = preprocessing.normalize(fc2.reshape((1,512)), norm='l2', axis=1)
diff = np.subtract(norm_fc1, norm_fc2)
dist = np.sqrt(np.sum(np.square(diff), 1))/2
print(dist)
for i in np.arange(20):
print(np.sqrt(np.sum(np.square(diff[0][i*25:i*25+25]), 0))/2)
fig = plt.figure()
ax = fig.add_subplot(2,1,1)
ax.plot(np.arange(512), norm_fc1[0])
# ax = fig.add_subplot(2,1,2)
ax.plot(np.arange(512), norm_fc2[0])
ax = fig.add_subplot(2,1,2)
ax.plot(np.arange(512), diff[0])
plt.show()
def main():
# with open('/home/anhdq23/Desktop/nguyen/VT_simulation/weights/arcface_ret50.json', 'r') as f:
# model_json = json.load(f)
# model = model_from_json(model_json)
# model.load_weights('/home/anhdq23/Desktop/nguyen/VT_simulation/weights/arcface_ret50.h5')
# model.summary()
cfg = load_yaml('./configs/arc_res50_new.yaml')
model = ArcFaceModel(size=cfg['input_size'],
backbone_type=cfg['backbone_type'],
num_classes=cfg['num_classes'],
head_type=cfg['head_type'],
embd_shape=cfg['embd_shape'],
w_decay=cfg['w_decay'],
training=False)
model.summary()
ckpt_path = tf.train.latest_checkpoint('./checkpoints/' + cfg['sub_name'])
print(ckpt_path)
if ckpt_path is not None:
print("[*] load ckpt from {}".format(ckpt_path))
model.load_weights(ckpt_path)
else:
print("[*] training from scratch.")
model_mask = ArcFaceModel(size=cfg['input_size'],
backbone_type=cfg['backbone_type'],
num_classes=cfg['num_classes'],
head_type=cfg['head_type'],
embd_shape=cfg['embd_shape'],
w_decay=cfg['w_decay'],
training=False)
ckpt_path = tf.train.latest_checkpoint('./checkpoints/' + 'arc_res50_mask')
print(ckpt_path)
if ckpt_path is not None:
print("[*] load ckpt from {}".format(ckpt_path))
model_mask.load_weights(ckpt_path)
else:
print("[*] training from scratch.")
import sys
sys.path.append('/home/anhdq23/Desktop/nguyen/VT_simulation/')
from detector import get_detector
predictor = get_detector()
ICPR_dict = dict()
path_ICPR = '/home/anhdq23/Desktop/nguyen/data/ICPR_cropped_face'
for name_fold in os.listdir(path_ICPR):
print(name_fold)
path_fold = os.path.join(path_ICPR, name_fold)
if name_fold not in ICPR_dict.keys():
ICPR_dict[name_fold] = []
for name_image in os.listdir(path_fold):
path_image = os.path.join(path_fold, name_image)
if '60' not in name_image[-10:-4] and '90' not in name_image[-10:-4]\
and '75' not in name_image[-10:]:
image = Image.open(path_image)
image = expand2square(image, (255, 255, 255))
image = image.resize((112, 112))
image = np.array(image)/255.0
_, labels, _ = predictor.predict(image, 1500/2, 0.6)
if labels.numpy()[0] == 1:
fc1 = model_mask.predict(image.reshape((1,112,112,3)))
norm_fc1 = preprocessing.normalize(fc1.reshape((1,cfg['embd_shape'])), norm='l2', axis=1)
else:
fc1 = model.predict(image.reshape((1,112,112,3)))
norm_fc1 = preprocessing.normalize(fc1.reshape((1,cfg['embd_shape'])), norm='l2', axis=1)
ICPR_dict[name_fold].append(norm_fc1)
path_ICPR = '/home/anhdq23/Desktop/nguyen/data/ICPR_cropped_face'
anchor_list = []
name_list = []
for name_fold in os.listdir(path_ICPR):
print(name_fold)
path_fold = os.path.join(path_ICPR, name_fold)
for name_image in os.listdir(path_fold):
path_image = os.path.join(path_fold, name_image)
if '+0+0' in name_image[-10:] or '+0-15' in name_image[-10:] or\
'+0+15' in name_image[-10:] or '+15+0' in name_image[-10:] or\
'-15+0' in name_image[-10:]:
print(name_image)
image = Image.open(path_image)
image = expand2square(image, (255, 255, 255))
image = image.resize((112, 112))
image = np.array(image)/255.0
_, labels, _ = predictor.predict(image, 1500/2, 0.6)
if labels.numpy()[0] == 1:
fc1 = model_mask.predict(image.reshape((1,112,112,3)))
norm_fc1 = preprocessing.normalize(fc1.reshape((1,cfg['embd_shape'])), norm='l2', axis=1)
else:
fc1 = model.predict(image.reshape((1,112,112,3)))
norm_fc1 = preprocessing.normalize(fc1.reshape((1,cfg['embd_shape'])), norm='l2', axis=1)
anchor_list.append(norm_fc1)
name_list.append(name_fold)
# Init faiss
import faiss
count_true = 0
count_all = 0
res = faiss.StandardGpuResources() # use a single GPU
index_flat = faiss.IndexFlatL2(512)
# gpu_index_flat = faiss.index_cpu_to_gpu(res, 0, index_flat)
gpu_index_flat = index_flat
gpu_index_flat.add(np.array(anchor_list).reshape((-1,512)))
for key in list(ICPR_dict.keys()):
for feature in ICPR_dict[key]:
D, I = gpu_index_flat.search(feature, k=1) # actual search
print(key, name_list[I[0][0]])
if key == name_list[I[0][0]]:
count_true +=1
count_all +=1
print(count_true, count_all)
def main(_):
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
os.environ['CUDA_VISIBLE_DEVICES'] = FLAGS.gpu
logger = tf.get_logger()
logger.disabled = True
logger.setLevel(logging.FATAL)
set_memory_growth()
cfg = load_yaml(FLAGS.cfg_path)
model = ArcFaceModel(size=cfg['input_size'],
backbone_type=cfg['backbone_type'],
num_classes=cfg['num_classes'],
head_type=cfg['head_type'],
embd_shape=cfg['embd_shape'],
w_decay=cfg['w_decay'],
training=True)
model.summary(line_length=80)
if cfg['train_dataset']:
logging.info("load ms1m dataset.")
dataset_len = cfg['num_samples']
steps_per_epoch = dataset_len // cfg['batch_size']
train_dataset = dataset.load_tfrecord_dataset(cfg['train_dataset'],
cfg['batch_size'],
cfg['binary_img'],
is_ccrop=cfg['is_ccrop'])
else:
logging.info("load fake dataset.")
dataset_len = 1
steps_per_epoch = 1
train_dataset = dataset.load_fake_dataset(cfg['input_size'])
learning_rate = tf.constant(cfg['base_lr'])
optimizer = tf.keras.optimizers.SGD(learning_rate=learning_rate,
momentum=0.9,
nesterov=True)
loss_fn = SoftmaxLoss()
ckpt_path = tf.train.latest_checkpoint('./checkpoints/' + cfg['sub_name'])
if ckpt_path is not None:
print("[*] load ckpt from {}".format(ckpt_path))
model.load_weights(ckpt_path)
epochs, steps = get_ckpt_inf(ckpt_path, steps_per_epoch)
else:
print("[*] training from scratch.")
epochs, steps = 1, 1
if FLAGS.mode == 'eager_tf':
# Eager mode is great for debugging
# Non eager graph mode is recommended for real training
summary_writer = tf.summary.create_file_writer('./logs/' +
cfg['sub_name'])
train_dataset = iter(train_dataset)
while epochs <= cfg['epochs']:
inputs, labels = next(train_dataset)
with tf.GradientTape() as tape:
logist = model(inputs, training=True)
reg_loss = tf.reduce_sum(model.losses)
pred_loss = loss_fn(labels, logist)
total_loss = pred_loss + reg_loss
grads = tape.gradient(total_loss, model.trainable_variables)
optimizer.apply_gradients(zip(grads, model.trainable_variables))
if steps % 5 == 0:
verb_str = "Epoch {}/{}: {}/{}, loss={:.2f}, lr={:.4f}"
print(
verb_str.format(epochs, cfg['epochs'],
steps % steps_per_epoch, steps_per_epoch,
total_loss.numpy(), learning_rate.numpy()))
with summary_writer.as_default():
tf.summary.scalar('loss/total loss',
total_loss,
step=steps)
tf.summary.scalar('loss/pred loss', pred_loss, step=steps)
tf.summary.scalar('loss/reg loss', reg_loss, step=steps)
tf.summary.scalar('learning rate',
optimizer.lr,
step=steps)
if steps % cfg['save_steps'] == 0:
print('[*] save ckpt file!')
model.save_weights('checkpoints/{}/e_{}_b_{}.ckpt'.format(
cfg['sub_name'], epochs, steps % steps_per_epoch))
steps += 1
epochs = steps // steps_per_epoch + 1
else:
model.compile(optimizer=optimizer,
loss=loss_fn,
run_eagerly=(FLAGS.mode == 'eager_fit'))
mc_callback = ModelCheckpoint(
'checkpoints/' + cfg['sub_name'] + '/e_{epoch}_b_{batch}.ckpt',
save_freq=cfg['save_steps'] * cfg['batch_size'],
verbose=1,
save_weights_only=True)
tb_callback = TensorBoard(log_dir='logs/',
update_freq=cfg['batch_size'] * 5,
profile_batch=0)
tb_callback._total_batches_seen = steps
tb_callback._samples_seen = steps * cfg['batch_size']
callbacks = [mc_callback, tb_callback]
history = model.fit(train_dataset,
epochs=cfg['epochs'],
steps_per_epoch=steps_per_epoch,
callbacks=callbacks,
initial_epoch=epochs - 1)
print("[*] training done!")
cfg = load_yaml(FLAGS.cfg_path)
cfg['num_classes'] = 179721
cfg['num_samples'] = 8621403
cfg['train_dataset'] = 'data/popet/dtat/ms1m_asian.tfrecord'
cfg['sub_name'] = '200324_model'
print(cfg)
model = ArcFaceModel(size=cfg['input_size'],
backbone_type=cfg['backbone_type'],
num_classes=cfg['num_classes'],
head_type=cfg['head_type'],
embd_shape=cfg['embd_shape'],
w_decay=cfg['w_decay'],
training=True)
model.summary(line_length=80)
if cfg['train_dataset']:
logging.info("load ms1m dataset.")
dataset_len = cfg['num_samples']
steps_per_epoch = dataset_len // cfg['batch_size']
train_dataset = dataset.load_tfrecord_dataset(
cfg['train_dataset'], cfg['batch_size'], cfg['binary_img'],
is_ccrop=cfg['is_ccrop'])
else:
logging.info("load fake dataset.")
dataset_len = 1
steps_per_epoch = 1
train_dataset = dataset.load_fake_dataset(cfg['input_size'])
learning_rate = tf.constant(cfg['base_lr'])
def main():
# with open('/home/anhdq23/Desktop/nguyen/VT_simulation/weights/arcface_ret50.json', 'r') as f:
# model_json = json.load(f)
# model = model_from_json(model_json)
# model.load_weights('/home/anhdq23/Desktop/nguyen/VT_simulation/weights/arcface_ret50.h5')
# model.summary()
cfg = load_yaml('./configs/arc_res50_mix.yaml')
model = ArcFaceModel(size=cfg['input_size'],
backbone_type=cfg['backbone_type'],
num_classes=cfg['num_classes'],
head_type=cfg['head_type'],
embd_shape=cfg['embd_shape'],
w_decay=cfg['w_decay'],
training=False)
model.summary()
ckpt_path = tf.train.latest_checkpoint('./checkpoints/' + cfg['sub_name'])
print(ckpt_path)
if ckpt_path is not None:
print("[*] load ckpt from {}".format(ckpt_path))
model.load_weights(ckpt_path)
else:
print("[*] training from scratch.")
# # serialize model to JSON
# model_json = model.to_json()
# with open("/home/anhdq23/Desktop/nguyen/image-segmentation-keras/weights/arc_res50_new.json", "w") as json_file:
# json.dump(model_json, json_file)
# model_mask.save_weights("/home/anhdq23/Desktop/nguyen/VT_simulation/weights/arc_res50_mask.h5")
data_path = '/home/anhdq23/Desktop/nguyen/arcface-tf2/data'
lfw, lfw_issame = get_val_pair(data_path, 'lfw_align_112/lfw')
lfw = np.transpose(lfw, [0, 2, 3, 1]) * 0.5 + 0.5
image_1 = lfw[0::2]
image_2 = lfw[1::2]
dist_all = []
for idx in range(len(lfw_issame)):
print(idx)
fc1 = model.predict(image_1[idx].reshape((1,112,112,3)))
norm_fc1 = preprocessing.normalize(fc1.reshape((1,cfg['embd_shape'])), norm='l2', axis=1)
fc2 = model.predict(image_2[idx].reshape((1,112,112,3)))
norm_fc2 = preprocessing.normalize(fc2.reshape((1,cfg['embd_shape'])), norm='l2', axis=1)
# dist = tf.keras.losses.cosine_similarity(fc1.reshape((1,512)), fc2.reshape((1,512)))
diff = np.subtract(norm_fc1, norm_fc2)
dist = np.sqrt(np.sum(np.square(diff), 1))/2
dist_all.extend(dist)
plt.plot(dist_all)
plt.show()
thresholds = np.arange(0, 1, 0.01)
tpr_all = []
fpr_all = []
for thr in thresholds:
tpr, fpr, acc, f1 = calculate_accuracy(thr, np.array(dist_all), lfw_issame)
top_left = np.sqrt((1-tpr)**2 + fpr**2)
print('thr %.4f' % thr , 'tpr %.4f' % tpr, 'fpr %.4f' % fpr, \
'top left %.4f' % top_left, 'acc %.4f' % acc, 'f1_score %.4f'%f1)
# top_left_batch.append(top_left)
tpr_all.append(tpr)
fpr_all.append(fpr)
for threshold in thresholds:
predict_issame = np.less(np.array(dist_all), threshold)
conf_matrix = confusion_matrix(lfw_issame, predict_issame)
print(conf_matrix)
plt.figure()
lw = 2
plt.plot(fpr_all, tpr_all, color='darkorange',
lw=lw, label='ROC curve')
plt.xlim([0.0, 1.])
plt.ylim([0.0, 1.])
plt.xlabel('False Positive Rate')
plt.ylabel('True Positive Rate')
plt.title('Receiver operating characteristic')
plt.legend(loc="lower right")
plt.show()
def main(_):
set_memory_growth()
cfg = load_yaml(FLAGS.cfg_path)
model = ArcFaceModel(size=cfg['input_size'],
backbone_type=cfg['backbone_type'],
num_classes=cfg['num_classes'],
head_type=cfg['head_type'],
embd_shape=cfg['embd_shape'],
w_decay=cfg['w_decay'],
training=True)
model.summary()
learning_rate = tf.constant(cfg['base_lr'])
optimizer = tf.keras.optimizers.SGD(learning_rate=learning_rate,
momentum=0.9,
nesterov=True)
loss_fn = SoftmaxLoss()
ckpt_path = tf.train.latest_checkpoint('./checkpoints/' + cfg['sub_name'])
if ckpt_path is not None:
print("[*] load ckpt from {}".format(ckpt_path))
model.load_weights(ckpt_path)
else:
print("[*] training from scratch.")
model.compile(optimizer=optimizer, loss=loss_fn)
data_path = 'data'
lfw, lfw_issame = get_val_pair(data_path, 'lfw_align_112/lfw')
lfw = np.transpose(lfw, [0, 2, 3, 1]) * 0.5 + 0.5
image_1 = lfw[0::2]
image_2 = lfw[1::2]
augment = ImgAugTransform()
if FLAGS.mode == 'eager_tf':
top_left_all = [0.008807]
for epochs in range(cfg['epochs']):
logging.info("Shuffle ms1m dataset.")
dataset_len = cfg['num_samples']
steps_per_epoch = dataset_len // cfg['batch_size']
train_dataset = dataset.load_tfrecord_dataset(
cfg['train_dataset'],
cfg['batch_size'],
cfg['binary_img'],
is_ccrop=cfg['is_ccrop'])
for batch, (x, y) in enumerate(train_dataset):
x0_new = np.array(x[0], dtype=np.uint8)
x1_new = np.array(x[1], dtype=np.float32)
for i in np.arange(len(x0_new)):
x0_new[i] = augment(x0_new[i])
temp = np.array(x0_new, dtype=np.float32) / 255.0
loss = model.train_on_batch(*((temp, x1_new), x1_new))
if batch % 50 == 0:
verb_str = "Epoch {}/{}: {}/{}, loss={:.6f}, lr={:.6f}"
print(
verb_str.format(
epochs, cfg['epochs'], batch, steps_per_epoch,
loss, cfg['base_lr'] / (1.0 + cfg['w_decay'] *
(epochs * 45489 + batch))))
if batch % cfg['save_steps'] == 0:
resnet_model = tf.keras.Model(
inputs=model.get_layer('resnet50').input,
outputs=model.get_layer('resnet50').output)
output_model = tf.keras.Model(
inputs=model.get_layer('OutputLayer').input,
outputs=model.get_layer('OutputLayer').output)
dist_all = []
top_left_batch = []
for idx in range(0, len(lfw_issame),
cfg['batch_size']):
tem = resnet_model.predict(
image_1[idx:idx + cfg['batch_size']])
embeds_1 = output_model.predict(tem)
norm_embeds_1 = preprocessing.normalize(embeds_1,
norm='l2',
axis=1)
tem = resnet_model.predict(
image_2[idx:idx + cfg['batch_size']])
embeds_2 = output_model.predict(tem)
norm_embeds_2 = preprocessing.normalize(embeds_2,
norm='l2',
axis=1)
diff = np.subtract(norm_embeds_1, norm_embeds_2)
dist = np.sqrt(np.sum(np.square(diff), 1)) / 2
dist_all.extend(dist)
thresholds = np.arange(0, 1, 0.01)
for thr in thresholds:
tpr, fpr, _ = calculate_accuracy(
thr, np.array(dist_all), lfw_issame)
top_left = np.sqrt((1 - tpr)**2 + fpr**2)
top_left_batch.append(top_left)
print(
"The current top left: {:.6f} Threshold: {:.2f}"
.format(np.min(top_left_batch),
0.01 * np.argmin(top_left_batch)))
if not len(top_left_all):
print(
"The best top left: {:.6f} Threshold: {:.2f}"
.format(np.min(top_left_batch),
0.01 * np.argmin(top_left_batch)))
else:
print("The best top left: {:.6f}".format(
top_left_all[-1]))
if not len(top_left_all):
top_left_all.append(np.min(top_left_batch))
print('[*] save ckpt file!')
model.save_weights(
'checkpoints/{}/e_{}_b_{}.ckpt'.format(
cfg['sub_name'], epochs,
batch % steps_per_epoch))
elif top_left_all[-1] > np.min(top_left_batch):
top_left_all.append(np.min(top_left_batch))
print('[*] save ckpt file!')
model.save_weights(
'checkpoints/{}/e_{}_b_{}.ckpt'.format(
cfg['sub_name'], epochs,
batch % steps_per_epoch))
model.save_weights('checkpoints/train_{}/{}.ckpt'.format(
cfg['sub_name'], cfg['sub_name']))