def calculate_goal_success(self, epoch_range):
if os.path.isfile('batch_succeed.txt'):
pass
else:
with open('batch_succeed.txt', 'w+') as f:
pass
sucess_rate_epoch, sucess_rate_epoch_std = [], []
error_in_dist = (self.error_in_dist**2 * 2)**0.5 # euclidean distance
for epoch in epoch_range:
dist_from_goal_per_agent = torch.tensor(Plot.extract_data(
epoch, dir=self.dir + os.sep, calculate_dist='ON'),
dtype=torch.float)
sucess_rate_batch = torch.sum(
dist_from_goal_per_agent <= error_in_dist,
dim=1) / dist_from_goal_per_agent.shape[1]
idx_of_batches_suceeded = (sucess_rate_batch == 1).nonzero()
with open('batch_succeed.txt', 'a+') as f:
f.write(
'In Epoch {0} the Total num of batches whos agent succeeded is: {1}\n'
.format(epoch, idx_of_batches_suceeded.shape[0]))
f.write('The batchs whos agents succeeded are:/n')
for batch in idx_of_batches_suceeded:
f.write(str(batch.item()))
f.write(',')
f.write('\n')
sucess_rate_batch = sucess_rate_batch.type(torch.FloatTensor)
sucess_rate_epoch += [torch.mean(sucess_rate_batch)]
sucess_rate_epoch_std += [torch.std(sucess_rate_batch)]
Plot.create_sucees_reate_plot(sucess_rate_epoch, sucess_rate_epoch_std,
epoch_range)
class CollectOutputAndTarget(Callback):
def __init__(self):
super(CollectOutputAndTarget, self).__init__()
self.inputs = [] # collect x_input batches
self.targets = [] # collect y_true batches
self.outputs = [] # collect y_pred batches
# the shape of these 2 variables will change according to batch shape
# to handle the "last batch", specify `validate_shape=False`
self.var_x_input = tf.Variable(0., validate_shape=False)
self.var_y_true = tf.Variable(0., validate_shape=False)
self.var_y_pred = tf.Variable(0., validate_shape=False)
self.train_plotObj = Plot('train', title='Train Predictions')
def on_batch_end(self, batch, logs=None):
# evaluate the variables and save them into lists
x_input = K.eval(self.var_x_input)
y_true = K.eval(self.var_y_true)
y_pred = K.eval(self.var_y_pred)
# print(y_true)
# print(y_pred)
# print(mat2uint8(y_true))
# print(mat2uint8(y_pred))
if showImages:
self.train_plotObj.showTrainResults(x_input[0, :, :, :],
y_true[0, :, :,
0], y_pred[0, :, :, 0])
class CollectOutputAndTarget(Callback):
def __init__(self):
super(CollectOutputAndTarget, self).__init__()
self.inputs = [] # collect x_input batches
self.targets = [] # collect y_true batches
self.outputs = [] # collect y_pred batches
# the shape of these 2 variables will change according to batch shape
# to handle the "last batch", specify `validate_shape=False`
self.var_x_input = tf.Variable(0., validate_shape=False)
self.var_y_true = tf.Variable(0., validate_shape=False)
self.var_y_pred = tf.Variable(0., validate_shape=False)
self.train_plotObj = Plot('train', title='Train Predictions')
def on_batch_end(self, batch, logs=None):
# evaluate the variables and save them into lists
x_input = K.eval(self.var_x_input)
y_true = K.eval(self.var_y_true)
y_pred = K.eval(self.var_y_pred)
# print(y_true)
# print(y_pred)
# print(mat2uint8(y_true))
# print(mat2uint8(y_pred))
if showImages:
self.train_plotObj.showTrainResults(x_input[0, :, :, :], y_true[0, :, :, 0], y_pred[0, :, :, 0])
def __init__(self):
super(CollectOutputAndTarget, self).__init__()
self.inputs = [] # collect x_input batches
self.targets = [] # collect y_true batches
self.outputs = [] # collect y_pred batches
# the shape of these 2 variables will change according to batch shape
# to handle the "last batch", specify `validate_shape=False`
self.var_x_input = tf.Variable(0., validate_shape=False)
self.var_y_true = tf.Variable(0., validate_shape=False)
self.var_y_pred = tf.Variable(0., validate_shape=False)
self.train_plotObj = Plot('train', title='Train Predictions')
def __init__(self):
super(CollectOutputAndTarget, self).__init__()
self.inputs = [] # collect x_input batches
self.targets = [] # collect y_true batches
self.outputs = [] # collect y_pred batches
# the shape of these 2 variables will change according to batch shape
# to handle the "last batch", specify `validate_shape=False`
self.var_x_input = tf.Variable(0., validate_shape=False)
self.var_y_true = tf.Variable(0., validate_shape=False)
self.var_y_pred = tf.Variable(0., validate_shape=False)
self.train_plotObj = Plot('train', title='Train Predictions')
def calculate(self, epoch_range, batch_range):
for epoch in epoch_range:
utterance = Plot.extract_data(epoch,
dir=self.dir,
calculate_utternace='ON')
total_iterations = utterance.shape[1]
for batch in batch_range:
for iteration in range(total_iterations):
self.counter += 1
cur_utter = utterance[batch, iteration, :, :]
for j in range(cur_utter.shape[0]):
for i in range(cur_utter.shape[1]):
if cur_utter[j, i] >= threshold:
self.utterance_stat[i] += 1
self.utterance_stat = np.true_divide(self.utterance_stat, self.counter)
print(self.utterance_stat)
def __init__(self, config, num_agents, num_landmarks, folder_dir):
super(GameModule, self).__init__()
self.batch_size = config.batch_size # scalar: num games in this batch
self.using_utterances = config.use_utterances # bool: whether current batch allows utterances
self.using_cuda = config.use_cuda
self.num_agents = num_agents # scalar: number of agents in this batch
self.num_landmarks = num_landmarks # scalar: number of landmarks in this batch
self.num_entities = self.num_agents + self.num_landmarks # type: int
self.world_dim = config.world_dim
self.time_horizon = config.time_horizon
self.num_epochs = config.num_epochs
self.folder_dir = folder_dir
if self.using_cuda:
self.Tensor = torch.cuda.FloatTensor
else:
self.Tensor = torch.FloatTensor
locations = torch.rand(self.batch_size, self.num_entities,
2) * config.world_dim
self.colors = (torch.rand(self.batch_size, self.num_entities, 1) *
config.num_colors).floor()
self.shapes = (torch.rand(self.batch_size, self.num_entities, 1) *
config.num_shapes).floor()
goal_agents = self.Tensor(self.batch_size, self.num_agents, 1)
goal_entities = (torch.rand(self.batch_size, self.num_agents, 1) *
self.num_landmarks).floor().long() + self.num_agents
goal_locations = self.Tensor(self.batch_size, self.num_agents, 2)
if self.using_cuda:
locations = locations.cuda()
self.colors = self.colors.cuda()
self.shapes = self.shapes.cuda()
goal_entities = goal_entities.cuda()
# [batch_size, num_entities, 2]
self.locations = Variable(locations)
# [batch_size, num_entities, 2]
self.physical = Variable(
torch.cat((self.colors, self.shapes), 2).float())
for b in range(self.batch_size):
goal_agents[b, :, 0] = torch.randperm(self.num_agents)
for b in range(self.batch_size):
goal_locations[b] = self.locations.data[b][
goal_entities[b].squeeze()]
# [batch_size, num_agents, 3]
self.goals = Variable(torch.cat((goal_locations, goal_agents), 2))
goal_agents = Variable(goal_agents)
if self.using_cuda:
self.memories = {
"physical":
Variable(
torch.zeros(self.batch_size, self.num_agents,
self.num_entities, config.memory_size).cuda()),
"action":
Variable(
torch.zeros(self.batch_size, self.num_agents,
config.memory_size).cuda())
}
else:
self.memories = {
"physical":
Variable(
torch.zeros(self.batch_size, self.num_agents,
self.num_entities, config.memory_size)),
"action":
Variable(
torch.zeros(self.batch_size, self.num_agents,
config.memory_size))
}
if self.using_utterances:
if self.using_cuda:
self.utterances = Variable(
torch.zeros(self.batch_size, self.num_agents,
config.vocab_size).cuda())
self.memories["utterance"] = Variable(
torch.zeros(self.batch_size, self.num_agents,
self.num_agents, config.memory_size).cuda())
else:
self.utterances = Variable(
torch.zeros(self.batch_size, self.num_agents,
config.vocab_size))
self.memories["utterance"] = Variable(
torch.zeros(self.batch_size, self.num_agents,
self.num_agents, config.memory_size))
agent_baselines = self.locations[:, :self.num_agents, :]
goals_by_landmark = torch.cat(
(goal_entities.type(torch.FloatTensor), goal_agents), 2).float()
sort_idxs = torch.sort(self.goals[:, :, 2])[1]
self.sorted_goals = Variable(self.Tensor(self.goals.size()))
# TODO: Bad for loop?
for b in range(self.batch_size):
self.sorted_goals[b] = self.goals[b][sort_idxs[b]]
self.sorted_goals = self.sorted_goals[:, :, :2]
# [batch_size, num_agents, num_entities, 2]
self.observations = self.locations.unsqueeze(
1) - agent_baselines.unsqueeze(2)
new_obs = self.goals[:, :, :2] - agent_baselines
# [batch_size, num_agents, 2] [batch_size, num_agents, 1]
self.observed_goals = torch.cat((new_obs, goal_agents), dim=2)
self.plots_matrix = Plot(self.batch_size, self.time_horizon,
self.num_entities, locations.shape[2],
self.world_dim, self.num_agents,
goals_by_landmark, self.folder_dir)
self.plots_matrix.save_plot_matrix("start", locations, self.colors,
self.shapes)
class GameModule(nn.Module):
def __init__(self, config, num_agents, num_landmarks, folder_dir):
super(GameModule, self).__init__()
self.batch_size = config.batch_size # scalar: num games in this batch
self.using_utterances = config.use_utterances # bool: whether current batch allows utterances
self.using_cuda = config.use_cuda
self.num_agents = num_agents # scalar: number of agents in this batch
self.num_landmarks = num_landmarks # scalar: number of landmarks in this batch
self.num_entities = self.num_agents + self.num_landmarks # type: int
self.world_dim = config.world_dim
self.time_horizon = config.time_horizon
self.num_epochs = config.num_epochs
self.folder_dir = folder_dir
if self.using_cuda:
self.Tensor = torch.cuda.FloatTensor
else:
self.Tensor = torch.FloatTensor
locations = torch.rand(self.batch_size, self.num_entities,
2) * config.world_dim
self.colors = (torch.rand(self.batch_size, self.num_entities, 1) *
config.num_colors).floor()
self.shapes = (torch.rand(self.batch_size, self.num_entities, 1) *
config.num_shapes).floor()
goal_agents = self.Tensor(self.batch_size, self.num_agents, 1)
goal_entities = (torch.rand(self.batch_size, self.num_agents, 1) *
self.num_landmarks).floor().long() + self.num_agents
goal_locations = self.Tensor(self.batch_size, self.num_agents, 2)
if self.using_cuda:
locations = locations.cuda()
self.colors = self.colors.cuda()
self.shapes = self.shapes.cuda()
goal_entities = goal_entities.cuda()
# [batch_size, num_entities, 2]
self.locations = Variable(locations)
# [batch_size, num_entities, 2]
self.physical = Variable(
torch.cat((self.colors, self.shapes), 2).float())
for b in range(self.batch_size):
goal_agents[b, :, 0] = torch.randperm(self.num_agents)
for b in range(self.batch_size):
goal_locations[b] = self.locations.data[b][
goal_entities[b].squeeze()]
# [batch_size, num_agents, 3]
self.goals = Variable(torch.cat((goal_locations, goal_agents), 2))
goal_agents = Variable(goal_agents)
if self.using_cuda:
self.memories = {
"physical":
Variable(
torch.zeros(self.batch_size, self.num_agents,
self.num_entities, config.memory_size).cuda()),
"action":
Variable(
torch.zeros(self.batch_size, self.num_agents,
config.memory_size).cuda())
}
else:
self.memories = {
"physical":
Variable(
torch.zeros(self.batch_size, self.num_agents,
self.num_entities, config.memory_size)),
"action":
Variable(
torch.zeros(self.batch_size, self.num_agents,
config.memory_size))
}
if self.using_utterances:
if self.using_cuda:
self.utterances = Variable(
torch.zeros(self.batch_size, self.num_agents,
config.vocab_size).cuda())
self.memories["utterance"] = Variable(
torch.zeros(self.batch_size, self.num_agents,
self.num_agents, config.memory_size).cuda())
else:
self.utterances = Variable(
torch.zeros(self.batch_size, self.num_agents,
config.vocab_size))
self.memories["utterance"] = Variable(
torch.zeros(self.batch_size, self.num_agents,
self.num_agents, config.memory_size))
agent_baselines = self.locations[:, :self.num_agents, :]
goals_by_landmark = torch.cat(
(goal_entities.type(torch.FloatTensor), goal_agents), 2).float()
sort_idxs = torch.sort(self.goals[:, :, 2])[1]
self.sorted_goals = Variable(self.Tensor(self.goals.size()))
# TODO: Bad for loop?
for b in range(self.batch_size):
self.sorted_goals[b] = self.goals[b][sort_idxs[b]]
self.sorted_goals = self.sorted_goals[:, :, :2]
# [batch_size, num_agents, num_entities, 2]
self.observations = self.locations.unsqueeze(
1) - agent_baselines.unsqueeze(2)
new_obs = self.goals[:, :, :2] - agent_baselines
# [batch_size, num_agents, 2] [batch_size, num_agents, 1]
self.observed_goals = torch.cat((new_obs, goal_agents), dim=2)
self.plots_matrix = Plot(self.batch_size, self.time_horizon,
self.num_entities, locations.shape[2],
self.world_dim, self.num_agents,
goals_by_landmark, self.folder_dir)
self.plots_matrix.save_plot_matrix("start", locations, self.colors,
self.shapes)
"""
Updates game state given all movements and utterances and returns accrued cost
- movements: [batch_size, num_agents, config.movement_size]
- utterances: [batch_size, num_agents, config.utterance_size]
- goal_predictions: [batch_size, num_agents, num_agents, config.goal_size]
Returns:
- scalar: total cost of all games in the batch
"""
def forward(self, movements, goal_predictions, utterances, t):
self.locations = self.locations + movements
self.plots_matrix.save_plot_matrix(t, self.locations, self.colors,
self.shapes) ####
agent_baselines = self.locations[:, :self.num_agents]
self.observations = self.locations.unsqueeze(
1) - agent_baselines.unsqueeze(2)
new_obs = self.goals[:, :, :2] - agent_baselines
goal_agents = self.goals[:, :, 2].unsqueeze(2)
self.observed_goals = torch.cat((new_obs, goal_agents), dim=2)
if self.using_utterances:
self.utterances = utterances
self.plots_matrix.save_utterance_matrix(utterances, t) ####
return self.compute_cost(movements, goal_predictions, utterances)
else:
return self.compute_cost(movements, goal_predictions)
def compute_cost(self, movements, goal_predictions, utterances=None):
physical_cost = self.compute_physical_cost()
movement_cost = self.compute_movement_cost(movements)
goal_pred_cost = self.compute_goal_pred_cost(goal_predictions)
return physical_cost + goal_pred_cost + movement_cost
"""
Computes the total cost agents get from being near their goals
agent locations are stored as [batch_size, num_agents + num_landmarks, entity_embed_size]
"""
def compute_physical_cost(self):
return 2 * torch.sum(
torch.sqrt(
torch.sum(
torch.pow(
self.locations[:, :self.num_agents, :] -
self.sorted_goals, 2), -1)))
"""
Computes the total cost agents get from predicting others' goals
goal_predictions: [batch_size, num_agents, num_agents, goal_size]
goal_predictions[., a_i, a_j, :] = a_i's prediction of a_j's goal with location relative to a_i
We want:
real_goal_locations[., a_i, a_j, :] = a_j's goal with location relative to a_i
We have:
goals[., a_j, :] = a_j's goal with absolute location
observed_goals[., a_j, :] = a_j's goal with location relative to a_j
Which means we want to build an observed_goals-like tensor but relative to each agent
real_goal_locations[., a_i, a_j, :] = goals[., a_j, :] - locations[a_i]
"""
def compute_goal_pred_cost(self, goal_predictions):
relative_goal_locs = self.goals.unsqueeze(
1)[:, :, :, :2] - self.locations.unsqueeze(
2)[:, :self.num_agents, :, :]
goal_agents = self.goals.unsqueeze(1)[:, :, :, 2:].expand_as(
relative_goal_locs)[:, :, :, -1:]
relative_goals = torch.cat((relative_goal_locs, goal_agents), dim=3)
return torch.sum(
torch.sqrt(
torch.sum(torch.pow(goal_predictions - relative_goals, 2),
-1)))
"""
Computes the total cost agents get from moving
"""
def compute_movement_cost(self, movements):
return torch.sum(torch.sqrt(torch.sum(torch.pow(movements, 2), -1)))
#dist, dist_per_agent = game.get_avg_agent_to_goal_distance() #add to tensorboard
def get_avg_agent_to_goal_distance(self):
dist_from_goal = self.locations[:, :self.
num_agents, :] - self.sorted_goals
euclidean_distance_per_batch = torch.sqrt(
torch.sum(torch.pow(dist_from_goal, 2), -1))
return torch.sum(
euclidean_distance_per_batch), euclidean_distance_per_batch
def test():
# Local Variables
num_test_images = None
# -----------------------------------------
# Network Testing Model - Importing Graph
# -----------------------------------------
# Loads the dataset and restores a specified trained model.
data = Dataloader()
# Searches dataset images filenames
if SELECTED_EVALUATION_SUBSET == 'test':
_, _, _, _, num_test_images = data.get_test_data(
test_split=args.test_split, test_file_path=args.test_file_path)
elif SELECTED_EVALUATION_SUBSET == 'train':
data.test_image_filenames, data.test_depth_filenames, _, _, num_test_images = data.get_train_data(
)
model = Test(data)
with tf.Session() as sess:
print('\n[Test] Loading the model...')
load_model(saver=tf.train.Saver(), sess=sess)
# ==============
# Testing Loop
# ==============
if args.debug:
num_test_images = 5 # Only for testing!
# TODO: Criar uma classe de test assim como fiz para train e valid, e declarar este objeto dentro dela
if args.show_test_results:
test_plot_obj = Plot(args.mode, title='Test Predictions')
print("[Test] Generating Predictions...")
pred_list, gt_list = [], []
timer = -time.time()
for i in tqdm(range(num_test_images)):
timer2 = -time.time()
# Evalute the network for the given image
# data.test_depth_filenames = [] # Only for testing the following condition!!! # FIXME: Atualmente, o código não dá suporte para esta situação
if data.test_depth_filenames: # It's not empty
feed_test = {
model.tf_image_key: data.test_image_filenames[i],
model.tf_depth_key: data.test_depth_filenames[i]
}
_, depth, depth_resized = sess.run(model.depth_op, feed_test)
else:
feed_test = {model.tf_image_key: data.test_image_filenames[i]}
_, image, image_resized, pred, pred_up = sess.run(
model.image_op + model.pred_op, feed_test)
# Clips Predictions at 50, 80 meters
try:
pred_50, pred_80 = sess.run(
[model.tf_pred_50, model.tf_pred_80], feed_test)
except AttributeError:
pred_50 = np.zeros((model.batch_size, ) +
model.output_size.get_size())
pred_80 = np.zeros((model.batch_size, ) +
model.output_size.get_size())
# Fill arrays for later on metrics evaluation
if args.eval_tool == 'monodepth':
pred_list.append(pred_up[0, :, :, 0])
gt_list.append(depth[:, :, 0])
# Saves the predictions and ground truth depth images as uint16 PNG Images
if SAVE_TEST_DISPARITIES or args.eval_tool == 'monodepth' or args.eval_tool == 'kitti_depth':
# Convert the Predictions Images from float32 to uint16
# print(data.test_image_filenames[i])
imsave_as_uint16_png(
settings.output_tmp_pred_dir + 'pred' + str(i) + '.png',
pred_up[0])
imsave_as_uint16_png(
settings.output_tmp_gt_dir + 'gt' + str(i) + '.png', depth)
timer2 += time.time()
# Show Results
if args.show_test_results:
# TODO: Fazer um lista the 'images_to_display' e dar append das imagens na lista
test_plot_obj.show_test_results(
image=image,
depth=depth[:, :, 0],
image_resized=image_resized,
depth_resized=depth_resized[:, :, 0],
pred=pred[0, :, :, 0],
pred_up=pred_up[0, :, :, 0],
pred_50=pred_50[0, :, :, 0],
pred_80=pred_80[0, :, :, 0],
i=i + 1)
# input("Continue...")
# Testing Finished.
timer += time.time()
print("Time elapsed: {} s\n".format(timer))
# =========
# Results
# =========
# Calculates Metrics
if args.eval_tool:
if data.test_depth_filenames:
print(
"[Test/Metrics] Calculating Metrics based on Test Predictions..."
)
print('args.test_split:', args.test_split)
print('args.test_file_path:', args.test_file_path)
print('dataset_path:', data.dataset.dataset_path)
print()
# Invokes Evaluation Tools
if args.eval_tool == 'monodepth':
pred_depths, gt_depths = metrics.generate_depth_maps(
pred_list, gt_list, data.dataset.dataset_path)
metrics.evaluation_tool_monodepth(pred_depths, gt_depths)
elif args.eval_tool == 'kitti_depth':
metrics.evaluation_tool_kitti_depth(num_test_images)
else:
print(
"[Test/Metrics] It's not possible to calculate metrics. There are no corresponding labels for generated predictions!"
)
else:
print("[Test/Metrics] Metrics calculation wasn't requested!")
def plot_metrics(directory_metrics, parts_directory_plot_metrics, legend):
print "\t\t[ Plotting metrics... ]"
for metric in metrics:
for i in range(0, 2):
d = directory_metrics + "/" + parts[i] + "/" + metric + "_" + parts[i]
print "\tPlotting %s ..." % d
if not os.path.isdir(d):
print "[%s directory not found]" % d
else:
xlog = False
ylog = False
if metric in metrics_log:
xlog = True
ylog = True
########################
# PLOT DISTRIBUTION
########################
data_distribution = {}
with open(d + "/" + metric + "_" + parts[i] + "_distribution.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_distribution[modules.extra.num(line[0])] = modules.extra.num(line[1])
title = ""
xlabel = metric
ylabel = ""
plot_file = parts_directory_plot_metrics[i] + "/" + xlabel + "_distribution"
plot_parameters = dict(marker="o", markersize=11, linewidth=0, alpha=0.5, color=colorset[0])
if data_distribution:
P = Plot(plot_file, title, xlabel, ylabel, xlog, ylog, [legend])
P.plot_single(data_distribution, plot_parameters)
else:
print "[Impossible to plot because no distribution data]"
########################
########################
# PLOT CDF
########################
data_reverse_cdf = {}
with open(d + "/" + metric + "_" + parts[i] + "_reverse_cdf.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_reverse_cdf[modules.extra.num(line[0])] = modules.extra.num(line[1])
title = ""
xlabel = metric
ylabel = ""
plot_file = parts_directory_plot_metrics[i] + "/" + xlabel + "_reverse_cdf"
plot_parameters = dict(linestyle="-", marker="s", markersize=8, linewidth=2.5, markeredgecolor=colorset[0], color=colorset[0])
if data_reverse_cdf:
P = Plot(plot_file, title, xlabel, ylabel, False, False, [legend], formatter=True)
P.plot_single(data_reverse_cdf, plot_parameters)
else:
print "[Impossible to plot because no reverse CDF data]"
def plot_correlations(directory_correlations, parts_directory_plot_correlations, legend):
print "\t\t[ Plotting correlations... ]"
for correlation in correlations:
metrics = correlation.split("-")
for i in range(0, 2):
correlation = metrics[0] + "_" + parts[i] + "-" + metrics[1] + "_" + parts[i]
d = directory_correlations + "/" + parts[i] + "/" + correlation
if not os.path.isdir(d):
print "[%s directory not found]" % d
else:
xlog = False
ylog = False
if metrics[0] in metrics_log:
xlog = True
if metrics[1] in metrics_log:
ylog = True
print "\tPlotting %s ..." % d
list_plot_parameters = []
list_data = []
########################
# SCATTER POINTS
########################
data_scatter = {}
with open(d + "/" + correlation + "_scatter.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_scatter[modules.extra.num(line[0])] = modules.extra.num(line[1])
highest = max(data_scatter.values())
list_plot_parameters.append(dict(marker="o", markersize=3, linewidth=0, alpha=0.5, color=colorset[0]))
list_data.append(data_scatter)
########################
# CURVE FITTING (AVG)
########################
data_avg_curve = {}
with open(d + "/" + correlation + "_avg_curve.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_avg_curve[modules.extra.num(line[0])] = modules.extra.num(line[1])
list_plot_parameters.append(dict(linestyle="--", marker="s", markersize=8, linewidth=2.5, color=colorset[1]))
list_data.append(data_avg_curve)
'''data_linear = {}
with open(d + "/" + correlation + "_linear_regression.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
if line[0] in nodes or not nodes:
data_linear[num(line[0])] = num(line[1])
list_plot_parameters.append(dict(linestyle="-", marker="o", markersize=6, linewidth=1.5, alpha=0.75))
list_data.append(data_linear)'''
'''data_fitting_curve = {}
with open(d + "/" + correlation + "_fitting_curve.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
if line[0] in nodes or not nodes:
data_fitting_curve[num(line[0])] = num(line[1])
list_plot_parameters.append(dict(linestyle="-", marker="o", markersize=6, linewidth=1.5, alpha=0.75))
list_data.append(data_fitting_curve)'''
title = ""
xlabel = metrics[0]
ylabel = metrics[1]
plot_file = parts_directory_plot_correlations[i] + "/" + correlation
try:
P = Plot(plot_file, title, xlabel, ylabel, True, True, [legend], ylimit=[0, highest])
P.plot_multiple(list_data, list_plot_parameters)
except Exception, e:
print "Error during plotting process : %s" % e
def plot_metrics(main_directory, list_directories, directory_dest_metrics, main_directory_paired = "", list_directories_paired = []):
print "\t\t[ Plotting metrics... ]"
for metric in metrics:
for j in range(0, 2):
m = metric + "_" + parts[j]
xlog = False
ylog = False
if metric in metrics_log:
xlog = True
ylog = True
print "\tPlotting %s ..." % m
xlabel = parts[j].title() + " " + metric
########################
# PLOT DISTRIBUTION
########################
list_plot_parameters = []
list_data = []
list_legend = []
i = 0
ci = 0
for directory in list_directories:
data_distribution = {}
with open(main_directory + "/" + directory + "/metrics/" + parts[j] + "/" + m + "/" + m + "_distribution.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_distribution[modules.extra.num(line[0])] = modules.extra.num(line[1])
list_data.append(data_distribution)
if paired:
list_plot_parameters.append(dict(marker=list_markers[i], markersize=12, linewidth=0, alpha=0.5, color=colorset_paired[ci]))
else:
list_plot_parameters.append(dict(marker=list_markers[i], markersize=12, linewidth=0, alpha=0.5, color=colorset[i]))
list_legend.append(directory)
if paired:
data_distribution = {}
with open(main_directory_paired + "/" + list_directories_paired[i] + "/metrics/" + parts[j] + "/" + m + "/" + m + "_distribution.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_distribution[modules.extra.num(line[0])] = modules.extra.num(line[1])
list_data.append(data_distribution)
list_plot_parameters.append(dict(marker=list_markers[i], markersize=12, linewidth=0, alpha=0.5, color=colorset_paired[ci + 1]))
list_legend.append(list_directories_paired[i])
i += 1
ci += 2
title = ""
ylabel = ""
plot_file = directory_dest_metrics + "/" + parts[j] + "/distribution_" + m
P = Plot(plot_file, title, xlabel, ylabel, xlog, ylog, list_legend)
P.plot_multiple(list_data, list_plot_parameters)
########################
########################
# PLOT CDF
########################
list_plot_parameters = []
list_data = []
list_legend = []
i = 0
ci = 0
for directory in list_directories:
data_reverse_cdf = {}
with open(main_directory + "/" + directory + "/metrics/" + parts[j] + "/" + m + "/" + m + "_reverse_cdf.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_reverse_cdf[modules.extra.num(line[0])] = modules.extra.num(line[1])
list_data.append(data_reverse_cdf)
if paired:
list_plot_parameters.append(dict(linestyle="-", marker=list_markers[i], markersize=14, linewidth=2.5, alpha=0.5, markeredgecolor=colorset_paired[ci], color=colorset_paired[ci]))
else:
list_plot_parameters.append(dict(linestyle="-", marker=list_markers[i], markersize=14, linewidth=2.5, alpha=0.5, markeredgecolor=colorset[i], color=colorset[i]))
list_legend.append(directory)
if paired:
data_reverse_cdf = {}
with open(main_directory_paired + "/" + list_directories_paired[i] + "/metrics/" + parts[j] + "/" + m + "/" + m + "_reverse_cdf.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_reverse_cdf[modules.extra.num(line[0])] = modules.extra.num(line[1])
list_data.append(data_reverse_cdf)
list_plot_parameters.append(dict(linestyle="-", marker=list_markers[i], markersize=14, linewidth=2.5, alpha=0.5, markeredgecolor=colorset_paired[ci + 1], color=colorset_paired[ci + 1]))
list_legend.append(list_directories_paired[i])
i += 1
ci += 2
title = ""
ylabel = ""
plot_file = directory_dest_metrics + "/" + parts[j] + "/reverse_cdf_" + m
P = Plot(plot_file, title, xlabel, ylabel, False, False, list_legend, formatter=True)
P.plot_multiple(list_data, list_plot_parameters)
def plot_correlations(main_directory, list_directories, directory_dest_correlations, main_directory_paired = "", list_directories_paired = []):
print "\t\t[ Plotting correlations... ]"
for correlation in correlations:
metrics = correlation.split("-")
for j in range(0, 2):
correlation = metrics[0] + "_" + parts[j] + "-" + metrics[1] + "_" + parts[j]
xlog = False
ylog = False
if metrics[0] in metrics_log:
xlog = True
if metrics[1] in metrics_log:
ylog = True
print "\tPlotting %s ..." % correlation
xlabel = parts[j].title() + " " + metrics[0]
ylabel = parts[j].title() + " " + metrics[1]
list_plot_parameters = []
list_data = []
list_data_avg_curve = []
list_plot_parameters_avg_curve = []
list_legend = []
highest = 0.0
i = 0
ci = 0
for directory in list_directories:
maximum = 0.0
data_scatter = {}
with open(main_directory + "/" + directory + "/correlations/" + parts[j] + "/" + correlation + "/" + correlation + "_scatter.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_scatter[modules.extra.num(line[0])] = modules.extra.num(line[1])
maximum = max(data_scatter.values())
if highest < maximum:
highest = maximum
if paired:
list_plot_parameters.append(dict(marker=list_markers[i], markersize=8, linewidth=0, alpha=0.5, color=colorset_paired[ci]))
else:
list_plot_parameters.append(dict(marker=list_markers[i], markersize=8, linewidth=0, alpha=0.5, color=colorset[i]))
list_data.append(data_scatter)
data_avg_curve = {}
with open(main_directory + "/" + directory + "/correlations/" + parts[j] + "/" + correlation + "/" + correlation + "_avg_curve.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_avg_curve[modules.extra.num(line[0])] = modules.extra.num(line[1])
if paired:
list_plot_parameters_avg_curve.append(dict(linestyle="--", marker=list_markers[i], markersize=13, linewidth=2.5, color=colorset_paired[ci]))
else:
list_plot_parameters_avg_curve.append(dict(linestyle="--", marker=list_markers[i], markersize=13, linewidth=2.5, color=colorset[i]))
if data_avg_curve:
list_data_avg_curve.append(data_avg_curve)
else:
print "[Impossible to retrieve average curve data from '%s']" % directory
list_legend.append(directory)
if paired:
maximum = 0.0
data_scatter = {}
with open(main_directory_paired + "/" + list_directories_paired[i] + "/correlations/" + parts[j] + "/" + correlation + "/" + correlation + "_scatter.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_scatter[modules.extra.num(line[0])] = modules.extra.num(line[1])
maximum = max(data_scatter.values())
if highest < maximum:
highest = maximum
list_plot_parameters.append(dict(marker=list_markers[i], markersize=8, linewidth=0, alpha=0.5, color=colorset_paired[ci + 1]))
list_data.append(data_scatter)
data_avg_curve = {}
with open(main_directory_paired + "/" + list_directories_paired[i] + "/correlations/" + parts[j] + "/" + correlation + "/" + correlation + "_avg_curve.data", 'r') as file:
for line in file:
line = line.replace("\n", "").split()
data_avg_curve[modules.extra.num(line[0])] = modules.extra(line[1])
list_plot_parameters_avg_curve.append(dict(linestyle="--", marker=list_markers[i], markersize=13, linewidth=2.5, color=colorset_paired[ci + 1]))
if data_avg_curve:
list_data_avg_curve.append(data_avg_curve)
else:
print "[Impossible to retrieve average curve data from %s]" % list_directories_paired[i]
list_legend.append(list_directories_paired[i])
i += 1
ci += 2
title = ""
plot_file = directory_dest_correlations + "/" + parts[j] + "/" + correlation
list_data.extend(list_data_avg_curve)
list_plot_parameters.extend(list_plot_parameters_avg_curve)
P = Plot(plot_file, title, xlabel, ylabel, xlog, ylog, list_legend, ylimit=[0, highest])
P.plot_multiple(list_data, list_plot_parameters)
def test(args):
print('[%s] Selected mode: Test' % appName)
# Local Variables
numSamples = None
# -----------------------------------------
# Network Testing Model - Importing Graph
# -----------------------------------------
# Loads the dataset and restores a specified trained model.
data = Dataloader(args)
# Searches dataset images filenames
if TEST_EVALUATE_SUBSET == 0:
_, _, _, _, numSamples = data.getTestData()
elif TEST_EVALUATE_SUBSET == 1:
data.test_image_filenames, data.test_depth_filenames, tf_test_image_filenames, tf_test_depth_filenames, numSamples = data.getTrainData(
)
model = Test(args, data)
with tf.Session() as sess:
print('\n[network/Testing] Loading the model...')
# Use to load from *.ckpt file
saver = tf.train.Saver()
saver.restore(sess, args.model_path)
# Use to load from npy file
# net.load(model_data_path, sess)
# ==============
# Testing Loop
# ==============
if args.show_test_results:
test_plotObj = Plot(args.mode, title='Test Predictions')
timer = -time.time()
pred_list, gt_list = [], []
for i in range(numSamples):
# for i in range(5): # Only for testing!
timer2 = -time.time()
# Evalute the network for the given image
# data.test_depth_filenames = [] # Only for testing the following condition!!! # FIXME: Atualmente, o código não dá suporte para esta situação
if data.test_depth_filenames: # It's not empty
feed_test = {
model.tf_image_key: data.test_image_filenames[i],
model.tf_depth_key: data.test_depth_filenames[i]
}
_, depth, depth_resized = sess.run(model.depth_op, feed_test)
else:
feed_test = {model.tf_image_key: data.test_image_filenames[i]}
_, image, image_resized = sess.run(model.image_op, feed_test)
pred, pred_up = sess.run(model.pred_op, feed_test)
# Clips Predictions at 50, 80 meters
try:
pred_50, pred_80 = sess.run(
[model.tf_pred_50, model.tf_pred_80], feed_test)
except AttributeError:
pred_50 = np.zeros((model.batch_size, ) +
model.output_size.getSize())
pred_80 = np.zeros((model.batch_size, ) +
model.output_size.getSize())
# Fill arrays for later on metrics evaluation
pred_list.append(pred_up[0, :, :, 0])
gt_list.append(depth[:, :, 0])
# Saves the Test Predictions as uint16 PNG Images
if SAVE_TEST_DISPARITIES:
# Convert the Predictions Images from float32 to uint16
pred_up_uint16 = exposure.rescale_intensity(pred_up[0],
out_range='float')
pred_up_uint16 = img_as_uint(pred_up_uint16)
depth_uint16 = exposure.rescale_intensity(depth,
out_range='float')
depth_uint16 = img_as_uint(depth_uint16)
# Save PNG Images
imageio.imsave("output/tmp/pred/pred" + str(i) + ".png",
pred_up_uint16)
imageio.imsave("output/tmp/gt/gt" + str(i) + ".png",
depth_uint16)
# Prints Testing Progress
timer2 += time.time()
print('step: %d/%d | t: %f | size(pred_list+gt_list): %d' %
(i + 1, numSamples, timer2,
total_size(pred_list) + total_size(gt_list)))
# break # Test
# Show Results
if args.show_test_results:
test_plotObj.showTestResults(image=image,
depth=depth[:, :, 0],
image_resized=image_resized,
depth_resized=depth_resized[:, :,
0],
pred=pred[0, :, :, 0],
pred_up=pred_up[0, :, :, 0],
pred_50=pred_50[0, :, :, 0],
pred_80=pred_80[0, :, :, 0],
i=i + 1)
# input("Continue...")
# Testing Finished.
timer += time.time()
print("\n[Network/Testing] Testing FINISHED! Time elapsed: %f s" %
timer)
# =========
# Results
# =========
# Calculate Metrics
if data.test_depth_filenames:
print(
"[Network/Testing] Calculating Metrics based on Testing Predictions..."
)
pred_array = np.array(pred_list)
gt_array = np.array(gt_list)
# LainaMetrics.evaluate(pred_array, gt_array) # FIXME:
# myMetrics.evaluate(pred_array, gt_array) # FIXME:
MonodepthMetrics.evaluate(pred_array, gt_array)
else:
print(
"[Network/Testing] It's not possible to calculate Metrics. There are no corresponding labels for Testing Predictions!"
)