def distorted_inputs_unit(roi_paths,
roi_expand_paths,
labels,
trainable,
roi_size=128,
roi_expand_size=256):
filenames = roi_paths
expandfilenames = roi_expand_paths
labels = labels
filename, expandfilename, label = tf.train.slice_input_producer(
[filenames, expandfilenames, labels], shuffle=False)
num_process_threads = 4
images_and_labels = []
for thread_id in range(num_process_threads):
image_buffer = tf.read_file(filename)
bbox = []
image = image_preprocessing(image_buffer,
bbox=bbox,
train=trainable,
thread_id=thread_id,
image_size=roi_size)
# image = tf.image.rgb_to_hsv(image)
expand_image_buffer = tf.read_file(expandfilename)
bbox = []
expandimage = image_preprocessing(expand_image_buffer,
bbox=bbox,
train=trainable,
thread_id=thread_id,
image_size=roi_expand_size)
images_and_labels.append([image, expandimage, label])
batch_image, batch_expand_image, batch_label = tf.train.batch_join(
images_and_labels,
batch_size=FLAGS.batch_size,
capacity=2 * num_process_threads * FLAGS.batch_size)
images = tf.cast(batch_image, tf.float32)
images = tf.reshape(images,
shape=[FLAGS.batch_size, roi_size, roi_size, 3])
expand_images = tf.cast(batch_expand_image, tf.float32)
# print expand_images
expand_images = tf.reshape(
expand_images,
shape=[FLAGS.batch_size, roi_expand_size, roi_expand_size, 3])
return images, expand_images, tf.reshape(batch_label, [FLAGS.batch_size])
def parse_tfrecord_inception(params,
record,
width,
height,
is_training=True,
use_summary=False):
'''
Parse the records saved using the tensorflow official inception data build
https://github.com/tensorflow/models
'''
image_buffer, label, bbox, label_text = parse_example_proto(record)
image = image_preprocessing(image_buffer,
bbox,
is_training,
width,
height,
use_summary=use_summary)
# [batch, height, width, channels] range(-1.0,1.0)
label_one_hot = tf.one_hot(tf.squeeze(label, axis=-1),
params['num_labels'],
dtype=image.dtype)
return image, label_one_hot
def distorted_inputs(data, image_dir):
#data = load_data(FLAGS.data_dir)
filenames = [image_dir + d['filename'] for d in data]
label_indexes = [d['label_index'] for d in data]
filename, label_index = tf.train.slice_input_producer(
[filenames, label_indexes], shuffle=True)
num_preprocess_threads = 4
images_and_labels = []
for thread_id in range(num_preprocess_threads):
image_buffer = tf.read_file(filename)
bbox = []
train = True
image = image_preprocessing(image_buffer, bbox, train, thread_id)
images_and_labels.append([image, label_index])
images, label_index_batch = tf.train.batch_join(
images_and_labels,
batch_size=FLAGS.batch_size,
capacity=2 * num_preprocess_threads * FLAGS.batch_size,
allow_smaller_final_batch=True)
height = FLAGS.input_size
width = FLAGS.input_size
depth = 3
images = tf.cast(images, tf.float32)
images = tf.reshape(images, shape=[FLAGS.batch_size, height, width, depth])
return images, tf.reshape(label_index_batch, [FLAGS.batch_size])
def distorted_inputs_unit(dataset, trainable, shuffle=True):
filenames = dataset.images_names
labels = dataset.labels
filename, label = tf.train.slice_input_producer([filenames, labels], shuffle=shuffle)
num_process_threads = 4
images_and_labels = []
for thread_id in range(num_process_threads):
image_buffer = tf.read_file(filename)
bbox = []
image = image_preprocessing(
image_buffer,
bbox=bbox,
train=trainable,
thread_id=thread_id
)
# image = tf.image.rgb_to_hsv(image)
images_and_labels.append([image, label])
batch_image, batch_label = tf.train.batch_join(
images_and_labels,
batch_size=net_config.BATCH_SIZE,
capacity=2*num_process_threads*net_config.BATCH_SIZE
)
height = net_config.IMAGE_W
width = net_config.IMAGE_H
depth = 3
images = tf.cast(batch_image, tf.float32)
images = tf.reshape(images, shape=[net_config.BATCH_SIZE, height, width, depth])
return images, tf.reshape(batch_label, [net_config.BATCH_SIZE])
def distorted_inputs():
data = load_data(FLAGS.data_dir)
filenames = [ d['filename'] for d in data ]
label_indexes = [ d['label_index'] for d in data ]
filename, label_index = tf.train.slice_input_producer([filenames, label_indexes], shuffle=True)
num_preprocess_threads = 4
images_and_labels = []
for thread_id in range(num_preprocess_threads):
image_buffer = tf.read_file(filename)
bbox = []
train = True
image = image_preprocessing(image_buffer, bbox, train, thread_id)
images_and_labels.append([image, label_index])
images, label_index_batch = tf.train.batch_join(
images_and_labels,
batch_size=FLAGS.batch_size,
capacity=2 * num_preprocess_threads * FLAGS.batch_size)
height = FLAGS.input_size
width = FLAGS.input_size
depth = 3
images = tf.cast(images, tf.float32)
images = tf.reshape(images, shape=[FLAGS.batch_size, height, width, depth])
return images, tf.reshape(label_index_batch, [FLAGS.batch_size])
def telemetry(sid, data):
if data:
# The current steering angle of the car
steering_angle = data["steering_angle"]
# The current throttle of the car
throttle = data["throttle"]
# The current speed of the car
speed = data["speed"]
# The current image from the center camera of the car
imgString = data["image"]
image = Image.open(BytesIO(base64.b64decode(imgString)))
image_array = np.asarray(image)
image_array = ip.image_preprocessing(image_array)
steering_angle = float(
model.predict(image_array[None, :, :, :], batch_size=1))
throttle = controller.update(float(speed))
print(steering_angle, throttle)
send_control(steering_angle, throttle)
# save frame
if args.image_folder != '':
timestamp = datetime.utcnow().strftime('%Y_%m_%d_%H_%M_%S_%f')[:-3]
image_filename = os.path.join(args.image_folder, timestamp)
image.save('{}.jpg'.format(image_filename))
else:
# NOTE: DON'T EDIT THIS.
sio.emit('manual', data={}, skip_sid=True)
def distorted_inputs(val_dataset):
# data = load_data(FLAGS.data_dir)
# filenames = [ d['filename'] for d in data ]
# label_indexes = [ d['label_index'] for d in data ]
# val_positive_path = '/home/give/Documents/dataset/BOT_Game/data_NY/patch/val/positive'
# val_negative_path = '/home/give/Documents/dataset/BOT_Game/data_NY/patch/val/negative'
filenames = val_dataset.images_names
label_indexes = val_dataset.labels
filename, label_index = tf.train.slice_input_producer(
[filenames, label_indexes], shuffle=False, num_epochs=1)
num_preprocess_threads = 1
images_and_labels = []
for thread_id in range(num_preprocess_threads):
image_buffer = tf.read_file(filename)
bbox = []
train = True
image = image_preprocessing(image_buffer, bbox, train, thread_id)
images_and_labels.append([image, label_index])
images, label_index_batch = tf.train.batch_join(
images_and_labels,
batch_size=FLAGS.batch_size,
allow_smaller_final_batch=True)
height = FLAGS.input_size
width = FLAGS.input_size
depth = 3
images = tf.cast(images, tf.float32)
images = tf.reshape(images, shape=[FLAGS.batch_size, height, width, depth])
return images, tf.reshape(label_index_batch, [FLAGS.batch_size])
def main(argv=None):
with tf.Session() as sess:
data_dir = FLAGS.data_dir
files = [os.path.join(data_dir, item) for item in os.listdir(data_dir)]
# files = random.sample(files, 800)
images = tf.placeholder(tf.float32,
[None, RESIZE_FINAL, RESIZE_FINAL, 3])
logits = inference(
images,
False,
num_classes=2,
num_blocks=[3, 4, 6, 3], # defaults to 50-layer network
use_bias=False, # defaults to using batch norm
bottleneck=True)
init = tf.global_variables_initializer()
resnet_variables = tf.global_variables()
saver = tf.train.Saver(resnet_variables)
saver.restore(sess, os.path.join(FLAGS.model_dir, FLAGS.ckpt_file))
softmax_output = tf.nn.softmax(logits)
if FLAGS.target:
print('Creating output file %s' % FLAGS.target)
output = open(os.path.join(FLAGS.data_dir, FLAGS.target), 'w')
writer = csv.writer(output)
writer.writerow(('file', 'label', 'score'))
num_batches = int(math.ceil(len(files)) / MAX_BATCH_SZ)
pg = ProgressBar(num_batches)
# try:
for j in range(num_batches):
start_offset = j * MAX_BATCH_SZ
end_offset = min((j + 1) * MAX_BATCH_SZ, len(files))
batch_image_files = files[start_offset:end_offset]
images_ = []
for file in batch_image_files:
print file
image_buffer = tf.read_file(file)
bbox = []
image = image_preprocessing(image_buffer, [], False)
images_.append(image)
image_batch = tf.stack(images_)
batch_results = sess.run(softmax_output,
feed_dict={images: image_batch.eval()})
batch_sz = batch_results.shape[0]
for i in range(batch_sz):
output_i = batch_results[i]
best_i = np.argmax(output_i)
best_choice = (label_list[best_i], output_i[best_i])
if writer is not None:
f = batch_image_files[i]
writer.writerow(
(f, best_choice[0], '%.2f' % best_choice[1]))
pg.update()
pg.done()
def distorted_inputs():
data = load_data(FLAGS.data_dir)
# load the file list and labels
filenames = [d['filename'] for d in data]
label_indexes = [d['label_index'] for d in data]
# slice_input_producer return tensors
filename, label_index = tf.train.slice_input_producer(
[filenames, label_indexes], shuffle=True)
# create multiple processes to read the data
num_preprocess_threads = 4
images_and_labels = []
for thread_id in range(num_preprocess_threads):
image_buffer = tf.read_file(filename)
bbox = []
train = True
image = image_preprocessing(image_buffer, bbox, train, thread_id)
images_and_labels.append([image, label_index])
# create batches
images, label_index_batch = tf.train.batch_join(
images_and_labels,
batch_size=FLAGS.batch_size,
capacity=2 * num_preprocess_threads * FLAGS.batch_size)
height = FLAGS.input_size
width = FLAGS.input_size
depth = 3
images = tf.cast(images, tf.float32)
images = tf.reshape(images, shape=[FLAGS.batch_size, height, width, depth])
return images, tf.reshape(label_index_batch, [FLAGS.batch_size])
