def main():
''' --- SELECT DEVICES --- '''
# Select either gpu or cpu
device = torch.device("cuda" if args.cuda else "cpu")
# Select among available GPUs
if args.cuda:
os.environ["CUDA_VISIBLE_DEVICES"] = ','.join(
str(x) for x in args.gpudevice)
''' --- INIT NETWORK MODEL --- '''
projdir = sys.path[0]
# Path for saving and loading the network.
saveloadpath = os.path.join(projdir, 'experiment\\checkpoints',
args.exp_name + '.pth')
Path(os.path.dirname(saveloadpath)).mkdir(exist_ok=True, parents=True)
# Load selected network model and put it to right device
if args.model_name == 'pointnet':
classifier = PointNetCls(dim=args.pointCoordDim,
num_class=len(args.categories),
feature_transform=args.feature_transform)
elif args.model_name == 'pointnet2':
classifier = PointNet2ClsMsg(dim=args.pointCoordDim,
num_class=len(args.categories))
else:
raise Exception(
'Argument "model_name" does not match existent networks')
classifier = classifier.to(device)
''' --- LOAD NETWORK IF EXISTS --- '''
if os.path.exists(saveloadpath):
print('Using pretrained model found...')
checkpoint = torch.load(saveloadpath)
start_epoch = checkpoint[
'epoch'] + 1 # Just becase make sure counting starts from 1, 2, ..., rather than 0, 1, ..., when print the information of start_epoch
iteration = checkpoint['iteration']
best_test_acc = checkpoint['test_accuracy']
classifier.load_state_dict(checkpoint['model_state_dict'])
else:
raise Exception('Model in: {} does not exists'.format(saveloadpath))
# put classifier in evaluation mode
classifier = classifier.eval()
''' --- INIT DATASETS AND DATALOADER (FOR SINGLE EPOCH) --- '''
# Ideal for PointNet and pointLSTM - dataloader will return (B:batch, S:seq, C:features, N:points)
dataTransformations = transforms.Compose([
ToSeries(),
DataAugmentation(),
Resampling(maxPointsPerFrame=10),
ToTensor()
])
# init dataset
nusc = NuScenes(version=args.nuscenes_eval_dir,
dataroot=args.nuscenes_dir,
verbose=True)
''' Iterate over samples '''
idx = list(range(len(nusc.sample)))
random.shuffle(idx) # shuffle samples
# for sample_rec in nusc.sample:
for i in idx:
sample_rec = nusc.sample[i]
sample_token = sample_rec['token']
# read all sensors data and merge them to a common reference frame
sensorlist = []
for sensor in args.sensors:
sensorlist.append(
MyRadarPointCloud.from_sample_token(nusc, sample_token,
sensor))
global radar_pc
radar_pc = MyRadarPointCloud.merge(sensorlist, 0)
# get annotations for actual sample
ann_tokens = nusc.get('sample', sample_token)['anns']
# filter annotations that are objects of interest for our classifier
def get_ann_properties(nusc, ann_token: str):
ann_rec = nusc.get('sample_annotation', ann_token)
category = ann_rec['category_name']
attr_tokens = ann_rec['attribute_tokens']
attr = [
nusc.get('attribute', attr_token)['name']
for attr_token in attr_tokens
]
if len(attr) == 0: attr = ['']
print('object category:{:30} attr:{:40} center:{:20}'.format(
category, str(attr),
str(ann_rec['translation'] - radar_pc.A_cs_2_gl[:3, 3])))
return category, attr[
0] # for now lets return only the first attribute (we might have more per object)
def isInterestingObject(ann_token):
cat, att = get_ann_properties(nusc, ann_token)
# category-attribute pair must match at least one of the desired category-attr pair
for desired_cat, desired_att in zip(args.categories,
args.attributes):
if cat in desired_cat and att in desired_att:
return True
return False
print('\n\nGround-truth objects on this frame:')
ann_tokens = list(filter(isInterestingObject, ann_tokens))
def getObjLabel(ann_token):
cat, att = get_ann_properties(nusc, ann_token)
for idx, (desired_cat, desired_att) in enumerate(
zip(args.categories, args.attributes)):
if cat in desired_cat and att in desired_att:
return idx
return np.NaN
labels = list(map(getObjLabel, ann_tokens))
assert np.all(
~np.isnan(labels)
), 'Something strange happened... object was selected as interesting but we can not find its label'
# create bounding boxes from annotations
ann_boxes = []
for ann_token, label in zip(ann_tokens, labels):
box = radar_pc.box(ann_token)
box.label = label
box.name = args.objectNames[label]
ann_boxes.append(box)
''' APPLY DBSCAN ON EACH SCENE'''
from sklearn.cluster import DBSCAN
points_scene = radar_pc.points # points_gl = <5,N>
# filter out objects, whose speed is less than 0.3 m/s
idx_moving = np.linalg.norm(points_scene[2:4, :].T, axis=1) > 0.25
points_scene = points_scene[:, idx_moving]
# apply DBSCAN
clustering = DBSCAN(eps=3, min_samples=1).fit(points_scene[:2, :].T)
''' For each cluster, run network and predict class '''
pred_results = []
for cluster_idx in range(max(clustering.labels_)):
# select points from the current cluster
points_idx = clustering.labels_ == cluster_idx
points_obj = points_scene[:, points_idx]
# apply necessary transformations
features = dataTransformations(points_obj)
# convert to torch.tensor
features = torch.tensor(features).type(
torch.FloatTensor).unsqueeze(0).to(device)
# calculate network prediction
pred = classifier(features).argmax().item()
# store result
pred_results.append((points_obj, pred))
''' RENDER '''
OBJCOLORS = ['red', 'green', 'blue', 'grey', 'orange', 'white']
fig = plt.figure(constrained_layout=True)
gs = GridSpec(2, 6, figure=fig)
gs.update(wspace=0, hspace=0)
axs = [
fig.add_subplot(gs[1, :3]),
fig.add_subplot(gs[1, 3:]),
fig.add_subplot(gs[0, 0]),
fig.add_subplot(gs[0, 1]),
fig.add_subplot(gs[0, 2]),
fig.add_subplot(gs[0, 3]),
fig.add_subplot(gs[0, 4]),
fig.add_subplot(gs[0, 5])
]
# render annotation boxes
for box in ann_boxes:
color = OBJCOLORS[box.label]
box.render(axs[0], colors=(color, ) * 3)
box.render(axs[1], colors=(color, ) * 3)
# render cars
radar_pc.car.render(axs[0], colors=('orange', 'k', 'k'))
radar_pc.car.render(axs[1], colors=('orange', 'k', 'k'))
# render pointcloud
radar_pc._render_pc(axs[0], radar_pc.points, color_channel='k')
for cluster_points, pred in pred_results:
# radar_pc._render_pc( axs[1], cluster_points, color_channel=OBJCOLORS[pred])
if pred != 5:
axs[1].scatter(cluster_points[0, :],
cluster_points[1, :],
s=30,
c=OBJCOLORS[pred],
edgecolors='k',
linewidths=1,
zorder=100)
# plot ellipses
if cluster_points.shape[1] > 1:
confidence_ellipse(axs[1],
cluster_points[0, :],
cluster_points[1, :],
edgecolor=OBJCOLORS[pred],
facecolor=OBJCOLORS[pred],
alpha=0.5)
# render camera images
cameras = [
'CAM_BACK_LEFT', 'CAM_FRONT_LEFT', 'CAM_FRONT', 'CAM_FRONT_RIGHT',
'CAM_BACK_RIGHT', 'CAM_BACK'
]
for ax_idx, cam_name in enumerate(cameras):
cam_token = sample_rec['data'][cam_name]
cam_rec = nusc.get('sample_data', cam_token)
img_filename = os.path.join(nusc.dataroot, cam_rec['filename'])
img = mpimg.imread(img_filename)
# axs[ax_idx].set_axis_off()
axs[ax_idx + 2].get_xaxis().set_visible(False)
axs[ax_idx + 2].get_yaxis().set_visible(False)
axs[ax_idx + 2].imshow(img)
axs[ax_idx + 2].set_title(cam_name)
# format
legend_elements = [
Line2D([0], [0], color=color, lw=4, label=name)
for color, name in zip(OBJCOLORS, args.objectNames)
]
axs[0].legend(handles=legend_elements,
loc='upper right',
prop={'size': 8})
axs[1].legend(handles=legend_elements,
loc='upper right',
prop={'size': 8})
axs[0].axis('equal')
axs[1].axis('equal')
mng = plt.get_current_fig_manager()
mng.window.showMaximized()
plt.tight_layout()
lims = (-60, 60)
axs[0].set_xlim(lims)
axs[0].set_ylim(lims)
axs[0].set_title(
'Radar detection points and ground truth bounding boxes of MOVING objects'
)
axs[1].set_xlim(lims)
axs[1].set_ylim(lims)
axs[1].set_title(
'Segmented segmentation of detection points (DBSCAN + PointNet classifier)\n3-sigma ellipses for each cluster and ground-truth bbs'
)
plt.tight_layout()
plt.show()
# plot scene with the standard Nuscenes method
# nusc.render_sample(sample_token)
# plt.show()
# sample_token = nusc.get('sample', sample_token)['next']
tb_writer.close()
def main(args):
''' --- SELECT DEVICES --- '''
# Select either gpu or cpu
device = torch.device("cuda" if args.cuda else "cpu")
# Select among available GPUs
if args.cuda:
os.environ["CUDA_VISIBLE_DEVICES"] = ','.join(
str(x) for x in args.gpudevice)
''' --- CREATE EXPERIMENTS DIRECTORY AND LOGGERS IN TENSORBOARD --- '''
projdir = sys.path[0]
# Path for saving and loading the network.
saveloadpath = os.path.join(projdir, 'experiment\\checkpoints',
args.exp_name + '.pth')
Path(os.path.dirname(saveloadpath)).mkdir(exist_ok=True, parents=True)
# timestamp = str(datetime.datetime.now().strftime('%Y-%m-%d-%H-%M'))
tblogdir = os.path.join(projdir, 'experiment\\tensorboardX',
args.exp_name) # + '_' + timestamp )
Path(tblogdir).mkdir(exist_ok=True, parents=True)
# Create tb_writer(the writer will be used to write the information on tb) by using SummaryWriter,
# flush_secs defines how much seconds need to wait for writing information.
tb_writer = SummaryWriter(logdir=tblogdir,
flush_secs=3,
write_to_disk=True)
''' --- INIT DATASETS AND DATALOADER (FOR SINGLE EPOCH) --- '''
# Read data from file, and create training data and testing data which are both in multiple frames. Beware Ts is
# recording for every frame, i.e. every 82ms the automotive radar records once to form single frame(We need this information for LSTM).
train_dataset, test_dataset, class_names = read_dataset(
args.datapath, Ts=0.082, train_test_split=0.8)
# Prepare the traing and testing dataset. both trainDataset and testDataset are dataset have multiple frames data,
# for each frame it contains the "unified number of detection points"(NMAX detection points per frame).
# Init test dataset(Beware we should NOT use data augmentation for test dataset)
test_dataTransformations = transforms.Compose(
[NormalizeTime(), Resampling(maxPointsPerFrame=10)])
testDataset = RadarClassDataset(dataset=test_dataset,
transforms=test_dataTransformations,
sequence_length=1)
# Init train datasets
train_dataTransformations = transforms.Compose([
NormalizeTime(),
DataAugmentation(),
Resampling(maxPointsPerFrame=10)
])
trainDataset = RadarClassDataset(dataset=train_dataset,
transforms=train_dataTransformations,
sequence_length=1)
# Create dataloader for training by using batch_size frames' data in each batch
trainDataLoader = DataLoader(trainDataset,
batch_size=args.batchsize,
shuffle=True,
num_workers=args.num_workers)
''' --- INIT NETWORK MODEL --- '''
# Load selected network model and put it to right device
if args.model_name == 'pointnet':
classifier = PointNetCls(dim=args.pointCoordDim,
num_class=args.numclasses,
feature_transform=args.feature_transform)
elif args.model_name == 'pointnet2':
classifier = PointNet2ClsMsg(
dim=args.pointCoordDim,
num_class=args.numclasses,
)
else:
raise Exception(
'Argument "model_name" does not match existent networks')
classifier = classifier.to(device)
''' --- LOAD NETWORK IF EXISTS --- '''
if os.path.exists(saveloadpath):
print('Using pretrained model found...')
checkpoint = torch.load(saveloadpath)
start_epoch = checkpoint[
'epoch'] + 1 # Just becase make sure counting starts from 1, 2, ..., rather than 0, 1, ..., when print the information of start_epoch
iteration = checkpoint['iteration']
best_test_acc = checkpoint['test_accuracy']
classifier.load_state_dict(checkpoint['model_state_dict'])
else:
print('No existing model, starting training from scratch...')
start_epoch = 1 # Just becase make sure counting starts from 1, 2, ..., rather than 0, 1, ..., when print the information of start_epoch
iteration = 1 # Just becase make sure counting starts from 1, 2, ..., rather than 0, 1, ..., when print the information of iteration
best_test_acc = 0
''' --- CREATE OPTIMIZER ---'''
if args.optimizer == 'SGD':
optimizer = torch.optim.SGD(classifier.parameters(),
lr=args.lr,
momentum=0.9)
elif args.optimizer == 'ADAM':
optimizer = torch.optim.Adam(classifier.parameters(),
lr=args.lr,
betas=(0.9, 0.999),
eps=1e-08,
weight_decay=args.decay_rate)
scheduler = torch.optim.lr_scheduler.StepLR(
optimizer, step_size=args.lr_epoch_half,
gamma=0.5) # half(0.5) the learning rate every 'step_size' epochs
# log info
printparams = 'Model parameters:' + json.dumps(
vars(args), indent=4, sort_keys=True)
print(printparams)
tb_writer.add_text('hyper-parameters', printparams,
iteration) # tb_writer.add_hparam(args)
tb_writer.add_text(
'dataset', 'dataset sample size: training: {}, test: {}'.format(
train_dataset.shape[0], test_dataset.shape[0]), iteration)
''' --- START TRANING ---'''
for epoch in range(start_epoch, args.epoch + 1):
print('Epoch %d/%s:' % (epoch, args.epoch))
# Add the "learning rate" into tensorboard scalar which will be shown in tensorboard
tb_writer.add_scalar('learning_rate', optimizer.param_groups[0]['lr'],
iteration)
for batch_id, data in tqdm(enumerate(trainDataLoader, 0),
total=len(trainDataLoader),
smoothing=0.9):
points, target = data # (B:batch x S:seq x C:features x N:points) , (B x S:seq)
# Squeeze to drop Sequence dimension, which is equal to 1, convert all the data to float(otherwise there will be data type problems when running the model) and move to device
points, target = points.squeeze(
dim=1).float().to(device), target.float().to(
device) # (B:batch x C:features x N:points) , (B)
# points, target = points.float().to(device), target.float().to(device)
# Reset gradients
optimizer.zero_grad()
# Sets the module in training mode
classifier = classifier.train()
# Forward propagation
pred = classifier(points)
# Calculate cross entropy loss (In the pointnet/pointnet2 network model, it outputs log_softmax result. Since
# "log_softmax -> nll_loss" == CrossEntropyLoss, so that we just need to call F.nll_loss)
loss = F.nll_loss(pred, target.long())
if args.model_name == 'pointnet':
loss += feature_transform_regularizer(classifier.trans) * 0.001
if args.feature_transform:
loss += feature_transform_regularizer(
classifier.trans_feat) * 0.001
# Back propagate
loss.backward()
# Update weights
optimizer.step()
# Log once for every 5 batches, add the "train_loss/cross_entropy" into tensorboard scalar which will be shown in tensorboard
if not batch_id % 5:
tb_writer.add_scalar('train_loss/cross_entropy', loss.item(),
iteration)
iteration += 1
# if batch_id> 2: break
scheduler.step()
''' --- TEST AND SAVE NETWORK --- '''
if not epoch % 10: # Doing the following things every epoch.
# Perform predictions on the training data.
train_targ, train_pred = test(classifier,
trainDataset,
device,
num_workers=args.num_workers,
batch_size=1800)
# Perform predictions on the testing data.
test_targ, test_pred = test(classifier,
testDataset,
device,
num_workers=args.num_workers,
batch_size=1800)
# Calculate the accuracy rate for training data.
train_acc = metrics_accuracy(train_targ, train_pred)
# Calculate the accuracy rate for testing data.
test_acc = metrics_accuracy(test_targ, test_pred)
print('\r Training loss: {}'.format(loss.item()))
print('Train Accuracy: {}\nTest Accuracy: {}'.format(
train_acc, test_acc))
# Add the "train_acc" "test_acc" into tensorboard scalars which will be shown in tensorboard.
tb_writer.add_scalars('metrics/accuracy', {
'train': train_acc,
'test': test_acc
}, iteration)
# Calculate confusion matrix.
confmatrix_test = metrics_confusion_matrix(test_targ, test_pred)
print('Test confusion matrix: \n', confmatrix_test)
# Log confusion matrix.
fig, ax = plot_confusion_matrix(confmatrix_test,
class_names,
normalize=False,
title='Test Confusion Matrix')
# Log normalized confusion matrix.
fig_n, ax_n = plot_confusion_matrix(
confmatrix_test,
class_names,
normalize=True,
title='Test Confusion Matrix - Normalized')
# Add the "confusion matrix" "normalized confusion matrix" into tensorboard figure which will be shown in tensorboard.
tb_writer.add_figure('test_confusion_matrix/abs',
fig,
global_step=iteration,
close=True)
tb_writer.add_figure('test_confusion_matrix/norm',
fig_n,
global_step=iteration,
close=True)
# Log precision recall curves.
for idx, clsname in enumerate(class_names):
# Convert log_softmax to softmax(which is actual probability) and select the desired class.
test_pred_binary = torch.exp(test_pred[:, idx])
test_targ_binary = test_targ.eq(idx)
# Add the "precision recall curves" which will be shown in tensorboard.
tb_writer.add_pr_curve(tag='pr_curves/' + clsname,
labels=test_targ_binary,
predictions=test_pred_binary,
global_step=iteration)
''' --- SAVE NETWORK --- '''
# if (test_acc >= best_test_acc): # For now lets save every time, since we are only testing in a subset of the test dataset
best_test_acc = test_acc # if test_acc > best_test_acc else best_test_acc
state = {
'epoch': epoch,
'iteration': iteration,
'train_accuracy': train_acc if args.train_metric else 0.0,
'test_accuracy': best_test_acc,
'model_state_dict': classifier.state_dict(),
'optimizer_state_dict': optimizer.state_dict(),
}
torch.save(state, saveloadpath)
print('Model saved!!!')
print('Best Accuracy: %f' % best_test_acc)
tb_writer.close()
def main():
''' --- SELECT DEVICES --- '''
# Select either gpu or cpu
device = torch.device("cuda" if args.cuda else "cpu")
# Select among available GPUs
if args.cuda: os.environ["CUDA_VISIBLE_DEVICES"] = ','.join(str(x) for x in args.gpudevice)
''' --- CREATE EXPERIMENTS DIRECTORY AND LOGGERS IN TENSORBOARD --- '''
projdir = sys.path[0]
# Path for saving and loading the network.
saveloadpath = os.path.join( projdir, 'experiment\\checkpoints', args.exp_name+'.pth')
Path(os.path.dirname(saveloadpath)).mkdir(exist_ok=True, parents=True)
# timestamp = str(datetime.datetime.now().strftime('%Y-%m-%d-%H-%M'))
tblogdir = os.path.join( projdir, 'experiment\\tensorboardX', args.exp_name ) # + '_' + timestamp )
Path(tblogdir).mkdir(exist_ok=True, parents=True)
# Create tb_writer(the writer will be used to write the information on tb) by using SummaryWriter,
# flush_secs defines how much seconds need to wait for writing information.
tb_writer = SummaryWriter( logdir=tblogdir, flush_secs=3, write_to_disk=True)
''' --- INIT DATASETS AND DATALOADER (FOR SINGLE EPOCH) --- '''
# Ideal for PointNet and pointLSTM - dataloader will return (B:batch, S:seq, C:features, N:points)
dataTransformations = transforms.Compose([
ToSeries(),
DataAugmentation(),
Resampling(maxPointsPerFrame=10),
ToTensor()
])
# Init nuScenes datasets
nusc_train = NuScenes(version=args.nuscenes_train_dir, dataroot=args.nuscenes_dir, verbose=True)
train_dataset = RadarClassDataset(nusc_train, categories=args.categories, sensors=args.sensors, transforms=dataTransformations, sequence_length=1)
nusc_test = NuScenes(version=args.nuscenes_test_dir, dataroot=args.nuscenes_dir, verbose=True)
test_dataset = RadarClassDataset(nusc_test, categories=args.categories, sensors=args.sensors, transforms=dataTransformations, sequence_length=1)
# Init training data loader
trainDataLoader = DataLoader(train_dataset, batch_size=args.batchsize, shuffle=True, num_workers=args.num_workers)
''' --- INIT NETWORK MODEL --- '''
# Load selected network model and put it to right device
if args.model_name == 'pointnet':
classifier = PointNetCls(dim=args.pointCoordDim, num_class=len(args.categories), feature_transform=args.feature_transform)
elif args.model_name == 'pointnet2':
classifier = PointNet2ClsMsg(dim=args.pointCoordDim, num_class=len(args.categories) )
else:
raise Exception('Argument "model_name" does not match existent networks')
classifier = classifier.to(device)
''' --- INIT LOSS FUNCTION --- '''
loss_fun = FocalLoss(gamma=args.focalLoss_gamma, num_classes=len(args.categories), alpha=args.weight_cat).to(device)
''' --- LOAD NETWORK IF EXISTS --- '''
if os.path.exists(saveloadpath):
print('Using pretrained model found...')
checkpoint = torch.load(saveloadpath)
start_epoch = checkpoint['epoch'] +1 # Just becase make sure counting starts from 1, 2, ..., rather than 0, 1, ..., when print the information of start_epoch
iteration = checkpoint['iteration']
best_test_acc = checkpoint['test_accuracy']
classifier.load_state_dict(checkpoint['model_state_dict'])
else:
print('No existing model, starting training from scratch...')
start_epoch = 1 # Just becase make sure counting starts from 1, 2, ..., rather than 0, 1, ..., when print the information of start_epoch
iteration = 1 # Just becase make sure counting starts from 1, 2, ..., rather than 0, 1, ..., when print the information of iteration
best_test_acc = 0
''' --- CREATE OPTIMIZER ---'''
if args.optimizer == 'SGD':
optimizer = torch.optim.SGD(
classifier.parameters(),
lr=args.lr,
momentum=0.9)
elif args.optimizer == 'ADAM':
optimizer = torch.optim.Adam(
classifier.parameters(),
lr=args.lr,
betas=(0.9, 0.999),
eps=1e-08,
weight_decay=args.decay_rate)
scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=args.lr_epoch_half, gamma=0.5) # half(0.5) the learning rate every 'step_size' epochs
# Log info
printparams = 'Model parameters:' + json.dumps(vars(args), indent=4, sort_keys=True)
print(printparams)
tb_writer.add_text('hyper-parameters',printparams,iteration) # tb_writer.add_hparam(args)
tb_writer.add_text('dataset','dataset sample size: training: {}, test: {}'.format(len(train_dataset),len(test_dataset)),iteration)
''' --- START TRANING ---'''
for epoch in range(start_epoch, args.epoch+1):
# epoch = start_epoch
print('Epoch %d/%s:' % (epoch, args.epoch))
# Add the "learning rate" into tensorboard scalar which will be shown in tensorboard
tb_writer.add_scalar('learning_rate', optimizer.param_groups[0]['lr'], iteration)
# Beware epochs_left = args.epoch - epoch
for batch_id, data in tqdm(enumerate(trainDataLoader, 0), total=len(trainDataLoader), smoothing=0.9):
points, target = data # (B:batch x S:seq x C:features x N:points) , (B x S:seq)
# Squeeze to drop Sequence dimension, which is equal to 1, convert all the data to float(otherwise there will be data type problems when running the model) and move to device
points, target = points.squeeze(dim=1).float().to(device), target.float().to(device) # (B:batch x C:features x N:points) , (B)
# points, target = points.float().to(device), target.float().to(device)
# Reset gradients
optimizer.zero_grad()
# Sets the module in training mode
classifier = classifier.train()
# Forward propagation
pred = classifier(points)
# MLE estimator = min (- log (softmax(x)) ) = min nll_loss(log_softmax(x))
# loss = F.nll_loss(pred, target.long())
loss = loss_fun(pred, target.long())
if args.model_name == 'pointnet':
loss += feature_transform_regularizer(classifier.trans) * 0.001
if args.feature_transform:
loss += feature_transform_regularizer(classifier.trans_feat) * 0.001
# Back propagate
loss.backward()
# Update weights
optimizer.step()
# Log once for every 5 batches, add the "train_loss/cross_entropy" into tensorboard scalar which will be shown in tensorboard
if not batch_id % 5: tb_writer.add_scalar('train_loss/cross_entropy', loss.item(), iteration)
iteration += 1
# Plot train confusion matrix every X steps
if not iteration % 20:
confmatrix_train = metrics_confusion_matrix(target, pred)
print('\nTrain confusion matrix: \n',confmatrix_train)
# We just finished one epoch
# if not batch_id+1 % int(train_dataset.len__()/args.batchsize):
''' --- TEST NETWORK --- '''
if not epoch % int(args.test_every_X_epochs): # Doing the following things every epoch.
# Perform predictions on the training data.
train_targ, train_pred = test(classifier, train_dataset, device, num_workers=0, batch_size=512)
# Perform predictions on the testing data.
test_targ, test_pred = test(classifier, test_dataset, device, num_workers=0, batch_size=512)
# Calculate the accuracy rate for training data.
train_acc = metrics_accuracy(train_targ, train_pred)
# Calculate the accuracy rate for testing data.
test_acc = metrics_accuracy(test_targ, test_pred)
print('\r Training loss: {}'.format(loss.item()))
print('Train Accuracy: {}\nTest Accuracy: {}'.format(train_acc, test_acc) )
# Add the "train_acc" "test_acc" into tensorboard scalars which will be shown in tensorboard.
tb_writer.add_scalars('metrics/accuracy', {'train':train_acc, 'test':test_acc}, iteration)
# Calculate confusion matrix.
confmatrix_test = metrics_confusion_matrix(test_targ, test_pred)
print('Test confusion matrix: \n',confmatrix_test)
# Log confusion matrix.
fig, ax = plot_confusion_matrix(confmatrix_test, args.categories, normalize=False, title='Test Confusion Matrix')
# Log normalized confusion matrix.
fig_n, ax_n = plot_confusion_matrix(confmatrix_test, args.categories, normalize=True, title='Test Confusion Matrix - Normalized')
# Add the "confusion matrix" "normalized confusion matrix" into tensorboard figure which will be shown in tensorboard.
tb_writer.add_figure('test_confusion_matrix/abs', fig, global_step=iteration, close=True)
tb_writer.add_figure('test_confusion_matrix/norm', fig_n, global_step=iteration, close=True)
# Log precision recall curves.
for idx, clsname in enumerate(args.categories):
# Convert log_softmax to softmax(which is actual probability) and select the desired class.
test_pred_binary = torch.exp(test_pred[:,idx])
test_targ_binary = test_targ.eq(idx)
# Add the "precision recall curves" which will be shown in tensorboard.
tb_writer.add_pr_curve(tag='pr_curves/'+clsname, labels=test_targ_binary, predictions=test_pred_binary, global_step=iteration)
# Store the best test accuracy
if (test_acc >= best_test_acc):
best_test_acc = max([best_test_acc, test_acc])
# NOTE: we possibly want to save only when when the best test accuracy is surpassed. For now lets save every X epoch
''' --- SAVE NETWORK --- '''
if not epoch % int(args.save_every_X_epochs):
print('Best Accuracy: %f'%best_test_acc)
state = {
'epoch': epoch,
'iteration': iteration,
'test_accuracy': best_test_acc,
'model_state_dict': classifier.state_dict(),
'optimizer_state_dict': optimizer.state_dict(),
}
torch.save(state, saveloadpath)
print('Model saved!!!')
# epoch += 1
# print('Epoch %d/%s:' % (epoch, args.epoch))
scheduler.step()
tb_writer.close()