def read_tensor_from_image_file(file_name,
input_height=299,
input_width=299,
input_mean=0,
input_std=255):
input_name = "file_reader"
output_name = "normalized"
file_reader = tf.read_file(file_name, input_name)
if file_name.endswith(".png"):
image_reader = tf.image.decode_png(file_reader,
channels=3,
name='png_reader')
elif file_name.endswith(".gif"):
image_reader = tf.squeeze(
tf.image.decode_gif(file_reader, name='gif_reader'))
elif file_name.endswith(".bmp"):
image_reader = tf.image.decode_bmp(file_reader, name='bmp_reader')
else:
image_reader = tf.image.decode_jpeg(file_reader,
channels=3,
name='jpeg_reader')
float_caster = tf.cast(image_reader, tf.float32)
dims_expander = tf.expand_dims(float_caster, 0)
resized = tf.image.resize_bilinear(dims_expander,
[input_height, input_width])
normalized = tf.divide(tf.subtract(resized, [input_mean]), [input_std])
sess = tf.Session()
result = sess.run(normalized)
return result
def visualize_tfrecords(path_to_tfrecord, num_vids, num_skip_frames):
"""Visualizes TFRecords in given path.
Args:
path_to_tfrecord: string, Path to TFRecords. Provide search pattern in
string.
num_vids: integer, Number of videos to visualize.
num_skip_frames: integer, Number of frames to skip while visualzing.
"""
tfrecord_files = glob.glob(path_to_tfrecord)
tfrecord_files.sort()
sess = tf.Session()
init_op = tf.group(tf.global_variables_initializer(),
tf.local_variables_initializer())
dataset = tf.data.TFRecordDataset(tfrecord_files)
dataset = dataset.map(decode)
dataset = dataset.batch(1)
iterator = dataset.make_one_shot_iterator()
next_batch = iterator.get_next()
sess.run(init_op)
for _ in xrange(num_vids):
# Fetch a new batch from the dataset
batch_videos, batch_names = sess.run(next_batch)
tf.logging.info('Class label = %d', batch_names[0])
for frame_idx in xrange(0, len(batch_videos[0]), num_skip_frames):
plt.imshow(batch_videos[0, frame_idx])
plt.pause(0.1)
plt.clf()
def start_bundle(self):
self._calc_loudness = tf.function(
_calc_loudness,
input_signature=[
tf.TensorSpec(shape=[_NUM_SECS * _AUDIO_RATE],
dtype=tf.float32)
])
self._sess = None if tf.executing_eagerly() else tf.Session()
def __call__(self):
with tf.Graph().as_default():
dataset = self._build_pipeline()
groundtruth = dataset.make_one_shot_iterator().get_next()
with tf.Session() as sess:
for _ in range(self._num_examples):
groundtruth_result = sess.run(groundtruth)
yield groundtruth_result
y = y.reshape(-1, 1)
#Convert the output labels to one-hot encoding format
y = np.eye(y.shape[1] + 1)[y]
y = np.array([item[0] for item in y])
X_train, X_test, y_train, y_test = model_selection.train_test_split(
X, y, test_size=.4, random_state=42)
num_outputs = y.shape[1]
num_inputs = X.shape[1]
learning_rate = 0.001
epochs = 3000
with tf.Session() as sess:
x = tf.placeholder(shape=[None, num_inputs], dtype=tf.float32)
y = tf.placeholder(shape=[None, num_outputs], dtype=tf.float32)
w = tf.Variable(tf.random.normal(shape=[num_inputs, num_outputs]),
dtype=tf.float32,
name='w')
b = tf.Variable(tf.random.normal(shape=[num_outputs]),
dtype=tf.float32,
name='b')
y_hat = tf.nn.sigmoid(tf.matmul(x, w) + b)
loss = tf.reduce_mean(-tf.reduce_sum(y * tf.log(y_hat), axis=1))
optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss)
match = tf.cast(tf.equal(tf.argmax(y, 1), tf.argmax(y_hat, 1)),
output_layer = args.output_layer
graph = load_graph(model_file)
t = read_tensor_from_image_file(file_name,
input_height=input_height,
input_width=input_width,
input_mean=input_mean,
input_std=input_std)
input_name = "import/" + input_layer
output_name = "import/" + output_layer
input_operation = graph.get_operation_by_name(input_name)
output_operation = graph.get_operation_by_name(output_name)
with tf.Session(graph=graph) as sess:
start = time.time()
results = sess.run(output_operation.outputs[0],
{input_operation.outputs[0]: t})
end = time.time()
results = np.squeeze(results)
top_k = results.argsort()[-5:][::-1]
labels = load_labels(label_file)
print('\nEvaluation time (1-image): {:.3f}s\n'.format(end - start))
for i in top_k:
if labels[i] == "bad canal images":
if results[i] >= 0.60:
shutil.copy2(singlefile, "result/bad")