def testwithdehaze():
model_dehaze = aodnet.AODnet()
model_dehaze.load_state_dict(
torch.load("./weight/dehaze.pth", map_location='cpu'))
if args.choose_net == "Unet":
model_segmentation = my_unet.UNet(3, 1)
elif args.choose_net == "Enet":
model_segmentation = enet.ENet(num_classes=1)
elif args.choose_net == "Segnet":
model_segmentation = segnet.SegNet(3, 1)
model_segmentation.load_state_dict(
torch.load(args.weight, map_location='cpu'))
liver_dataset = LiverDataset_three("data/val_dehaze",
transform=x_transform,
target_transform=y_transform)
dataloaders = DataLoader(liver_dataset) # batch_size默认为1
model_dehaze.eval()
model_segmentation.eval()
metric = SegmentationMetric(2)
# import matplotlib.pyplot as plt
# plt.ion()
mean_acc, mean_miou = [], []
with torch.no_grad():
for src, rain, mask in dataloaders:
y = model_dehaze(src)
y1 = model_segmentation(y)
y1 = torch.squeeze(y1).numpy()
y_label = torch.squeeze(mask).numpy()
y_label = y_label * 255
y1 = y1 * 127.5
# print(y_label.shape,y.shape)
image = np.concatenate((y_label, y1))
if args.choose_net == "Unet":
img_y = (y1 > 0.5)
elif args.choose_net == "Enet":
img_y = (y1 > 0.5)
elif args.choose_net == "Segnet":
img_y = (y1 > 0.5)
elif args.choose_net == "Scnn":
img_y = (y1 > 0.5)
img_y = img_y.astype(int)
y_label = y_label.astype(int)
metric.addBatch(img_y, y_label)
acc = metric.pixelAccuracy()
mIoU = metric.meanIntersectionOverUnion()
# confusionMatrix=metric.genConfusionMatrix(img_y, y_label)
mean_acc.append(acc)
mean_miou.append(mIoU)
# print(acc, mIoU,confusionMatrix)
print(acc, mIoU)
# plt.imshow(image)
# plt.pause(0.01)
# plt.show()
print("average acc:%0.6f average miou:%0.6f" %
(np.mean(mean_acc), np.mean(mean_miou)))
def test_save_loss_graphs_no_class_weight(self):
x = np.random.uniform(-1, 1, self.x_shape)
x = Variable(x.astype(np.float32))
t = np.random.randint(
0, 12, (self.x_shape[0], self.x_shape[2], self.x_shape[3]))
t = Variable(t.astype(np.int32))
for depth in six.moves.range(1, self.n_encdec + 1):
model = segnet.SegNet(n_encdec=self.n_encdec,
n_classes=12,
in_channel=self.x_shape[1])
model = segnet.SegNetLoss(model,
class_weight=None,
train_depth=depth)
y = model(x, t)
cg = build_computational_graph([y],
variable_style=_var_style,
function_style=_func_style).dump()
for e in range(1, self.n_encdec + 1):
self.assertTrue(
'encdec{}'.format(e) in model.predictor._children)
fn = 'tests/SegNet_xt_depth-{}_{}.dot'.format(self.n_encdec, depth)
if os.path.exists(fn):
continue
with open(fn, 'w') as f:
f.write(cg)
subprocess.call('dot -Tpng {} -o {}'.format(
fn, fn.replace('.dot', '.png')),
shell=True)
def main():
nClasses = args.nClasses
train_batch_size = 16
val_batch_size = 16
epochs = 50
img_height = 256
img_width = 256
root_path = '../../datasets/segmentation/'
mode = 'seg' if nClasses == 2 else 'parse'
train_file = './data/{}_train.txt'.format(mode)
val_file = './data/{}_test.txt'.format(mode)
if args.model == 'unet':
model = unet.Unet(nClasses,
input_height=img_height,
input_width=img_width)
elif args.model == 'segnet':
model = segnet.SegNet(nClasses,
input_height=img_height,
input_width=img_width)
else:
raise ValueError(
'Does not support {}, only supports unet and segnet now'.format(
args.model))
model.compile(loss='categorical_crossentropy',
optimizer=Adam(lr=1e-4),
metrics=['accuracy'])
model.summary()
train = segdata_generator.generator(root_path, train_file,
train_batch_size, nClasses, img_height,
img_width)
val = segdata_generator.generator(root_path,
val_file,
val_batch_size,
nClasses,
img_height,
img_width,
train=False)
if not os.path.exists('./results/'):
os.mkdir('./results')
save_file = './weights/{}_seg_weights.h5'.format(args.model) if nClasses == 2 \
else './weights/{}_parse_weights.h5'.format(args.model)
checkpoint = ModelCheckpoint(save_file,
monitor='val_acc',
save_best_only=True,
save_weights_only=True,
verbose=1)
history = model.fit_generator(
train,
steps_per_epoch=12706 // train_batch_size,
validation_data=val,
validation_steps=5000 // val_batch_size,
epochs=epochs,
callbacks=[checkpoint],
)
plot_history(history, './results/', args.model)
save_history(history, './results/', args.model)
def trainwithdehaze():
model_dehaze = aodnet.AODnet().to(device)
dsize = (3, 1, 256, 256)
# inputs1 = torch.randn(dsize).to(device)
# total_ops, total_params = profile(model_dehaze, (inputs1,), verbose=False)
# print(" %.2f | %.2f" % (total_params / (1000 ** 2), total_ops / (1000 ** 3)))
if args.choose_net == "Unet":
model_segmentation = my_unet.UNet(3, 1).to(device)
elif args.choose_net == "Enet":
model_segmentation = enet.ENet(num_classes=1).to(device)
elif args.choose_net == "Segnet":
model_segmentation = segnet.SegNet(3, 1).to(device)
# inputs2 = torch.randn(dsize).to(device)
# total_ops, total_params = profile(model_segmentation, (inputs2,), verbose=False)
# print(" %.2f | %.2f" % (total_params / (1000 ** 2), total_ops / (1000 ** 3)))
batch_size = args.batch_size
# dehaze的损失函数
criterion_dehaze = torch.nn.MSELoss()
# dehaze的优化函数
optimizer_dehaze = optim.Adam(model_dehaze.parameters(
)) # model.parameters():Returns an iterator over module parameters
# 语义分割的损失函数
criterion_segmentation = torch.nn.BCELoss()
# 语义分割的优化函数
optimizer_segmentation = optim.Adam(model_segmentation.parameters(
)) # model.parameters():Returns an iterator over module parameters
# 加载数据集
dataset_dehaze = LiverDataset_three("data/train_dehaze/",
transform=x_transform,
target_transform=y_transform)
dataloader_dehaze = DataLoader(dataset_dehaze,
batch_size=batch_size,
shuffle=True,
num_workers=4)
# DataLoader:该接口主要用来将自定义的数据读取接口的输出或者PyTorch已有的数据读取接口的输入按照batch size封装成Tensor
# batch_size：how many samples per minibatch to load，这里为4，数据集大小400，所以一共有100个minibatch
# shuffle:每个epoch将数据打乱，这里epoch=10。一般在训练数据中会采用
# num_workers：表示通过多个进程来导入数据，可以加快数据导入速度
train_dehaze_model(model_dehaze,
model_segmentation,
criterion_dehaze,
criterion_segmentation,
optimizer_dehaze,
optimizer_segmentation,
dataloader_dehaze,
num_epochs=6)
def load_model(model_name, noc):
if model_name == 'fcn':
model = fcn.FCN8s(noc)
if model_name == 'segnet':
model = segnet.SegNet(3, noc)
if model_name == 'pspnet':
model = pspnet.PSPNet(noc)
if model_name == 'unet':
model = unet.UNet(noc)
if model_name == 'segfast':
model = segfast.SegFast(64, noc)
if model_name == 'segfast_basic':
model = segfast_basic.SegFast_Basic(64, noc)
if model_name == 'segfast_mobile':
model = segfast_mobile.SegFast_Mobile(noc)
if model_name == 'segfast_v2_3':
model = segfast_v2.SegFast_V2(64, noc, 3)
if model_name == 'segfast_v2_5':
model = segfast_v2.SegFast_V2(64, noc, 5)
return model
def predict_segmentation():
n_classes = 2
images_path = '/home/deep/datasets/'
val_file = './data/seg_test.txt'
input_height = 256
input_width = 256
if args.model == 'unet':
m = unet.Unet(n_classes,
input_height=input_height,
input_width=input_width)
elif args.model == 'segnet':
m = segnet.SegNet(n_classes,
input_height=input_height,
input_width=input_width)
else:
raise ValueError('Do not support {}'.format(args.model))
m.load_weights("./results/{}_weights.h5".format(args.model))
m.compile(loss='categorical_crossentropy',
optimizer='adam',
metrics=['accuracy'])
colors = np.array([[0, 0, 0], [255, 255, 255]])
i = 0
for x, y in generator(images_path, val_file, 1, n_classes, input_height,
input_width):
pr = m.predict(x)[0]
pr = pr.reshape((input_height, input_width, n_classes)).argmax(axis=2)
seg_img = np.zeros((input_height, input_width, 3))
for c in range(n_classes):
seg_img[:, :,
0] += ((pr[:, :] == c) * (colors[c][0])).astype('uint8')
seg_img[:, :,
1] += ((pr[:, :] == c) * (colors[c][1])).astype('uint8')
seg_img[:, :,
2] += ((pr[:, :] == c) * (colors[c][2])).astype('uint8')
cv2.imshow('test', seg_img)
cv2.imwrite('./output/{}.jpg'.format(i), seg_img)
i += 1
cv2.waitKey(30)
def test_save_normal_graphs(self):
x = np.random.uniform(-1, 1, self.x_shape)
x = Variable(x.astype(np.float32))
for depth in six.moves.range(1, self.n_encdec + 1):
model = segnet.SegNet(n_encdec=self.n_encdec,
in_channel=self.x_shape[1])
y = model(x, depth)
cg = build_computational_graph([y],
variable_style=_var_style,
function_style=_func_style).dump()
for e in range(1, self.n_encdec + 1):
self.assertTrue('encdec{}'.format(e) in model._children)
fn = 'tests/SegNet_x_depth-{}_{}.dot'.format(self.n_encdec, depth)
if os.path.exists(fn):
continue
with open(fn, 'w') as f:
f.write(cg)
subprocess.call('dot -Tpng {} -o {}'.format(
fn, fn.replace('.dot', '.png')),
shell=True)
def test_remove_link(self):
opt = optimizers.MomentumSGD(lr=0.01)
# Update each depth
for depth in six.moves.range(1, self.n_encdec + 1):
model = segnet.SegNet(self.n_encdec, self.n_classes,
self.x_shape[1], self.n_mid)
model = segnet.SegNetLoss(model,
class_weight=None,
train_depth=depth)
opt.setup(model)
# Deregister non-target links from opt
if depth > 1:
model.predictor.remove_link('conv_cls')
for d in range(1, self.n_encdec + 1):
if d != depth:
model.predictor.remove_link('encdec{}'.format(d))
for name, link in model.namedparams():
if depth > 1:
self.assertTrue('encdec{}'.format(depth) in name)
else:
self.assertTrue('encdec{}'.format(depth) in name
or 'conv_cls' in name)
def test():
if args.choose_net == "Unet":
model = my_unet.UNet(3, 1).to(device)
if args.choose_net == "My_Unet":
model = my_unet.My_Unet2(3, 1).to(device)
elif args.choose_net == "Enet":
model = enet.ENet(num_classes=13).to(device)
elif args.choose_net == "Segnet":
model = segnet.SegNet(3, 1).to(device)
elif args.choose_net == "CascadNet":
model = my_cascadenet.CascadeNet(3, 1).to(device)
elif args.choose_net == "my_drsnet_A":
model = my_drsnet.MultiscaleSENetA(in_ch=3, out_ch=1).to(device)
elif args.choose_net == "my_drsnet_B":
model = my_drsnet.MultiscaleSENetB(in_ch=3, out_ch=1).to(device)
elif args.choose_net == "my_drsnet_C":
model = my_drsnet.MultiscaleSENetC(in_ch=3, out_ch=1).to(device)
elif args.choose_net == "my_drsnet_A_direct_skip":
model = my_drsnet.MultiscaleSENetA_direct_skip(in_ch=3,
out_ch=1).to(device)
elif args.choose_net == "SEResNet":
model = my_drsnet.SEResNet18(in_ch=3, out_ch=1).to(device)
elif args.choose_net == "resnext_unet":
model = resnext_unet.resnext50(in_ch=3, out_ch=1).to(device)
elif args.choose_net == "resnet50_unet":
model = resnet50_unet.UNetWithResnet50Encoder(in_ch=3,
out_ch=1).to(device)
elif args.choose_net == "unet_res34":
model = unet_res34.Resnet_Unet(in_ch=3, out_ch=1).to(device)
elif args.choose_net == "dfanet":
ch_cfg = [[8, 48, 96], [240, 144, 288], [240, 144, 288]]
model = dfanet.DFANet(ch_cfg, 3, 1).to(device)
elif args.choose_net == "cgnet":
model = cgnet.Context_Guided_Network(1).to(device)
elif args.choose_net == "lednet":
model = lednet.Net(num_classes=1).to(device)
elif args.choose_net == "bisenet":
model = bisenet.BiSeNet(1, 'resnet18').to(device)
elif args.choose_net == "espnet":
model = espnet.ESPNet(classes=1).to(device)
elif args.choose_net == "pspnet":
model = pspnet.PSPNet(1).to(device)
elif args.choose_net == "fddwnet":
model = fddwnet.Net(classes=1).to(device)
elif args.choose_net == "contextnet":
model = contextnet.ContextNet(classes=1).to(device)
elif args.choose_net == "linknet":
model = linknet.LinkNet(classes=1).to(device)
elif args.choose_net == "edanet":
model = edanet.EDANet(classes=1).to(device)
elif args.choose_net == "erfnet":
model = erfnet.ERFNet(classes=1).to(device)
dsize = (1, 3, 128, 192)
inputs = torch.randn(dsize).to(device)
total_ops, total_params = profile(model, (inputs, ), verbose=False)
print(" %.2f | %.2f" % (total_params / (1000**2), total_ops / (1000**3)))
model.load_state_dict(torch.load(args.weight))
liver_dataset = LiverDataset("data/val_camvid",
transform=x_transform,
target_transform=y_transform)
dataloaders = DataLoader(liver_dataset) # batch_size默认为1
model.eval()
metric = SegmentationMetric(13)
# import matplotlib.pyplot as plt
# plt.ion()
multiclass = 1
mean_acc, mean_miou = [], []
alltime = 0.0
with torch.no_grad():
for x, y_label in dataloaders:
x = x.to(device)
start = time.time()
y = model(x)
usingtime = time.time() - start
alltime = alltime + usingtime
if multiclass == 1:
# predict输出处理：
# https://www.cnblogs.com/ljwgis/p/12313047.html
y = F.sigmoid(y)
y = y.cpu()
# y = torch.squeeze(y).numpy()
y = torch.argmax(y.squeeze(0), dim=0).data.numpy()
print(y.max(), y.min())
# y_label = y_label[0]
y_label = torch.squeeze(y_label).numpy()
else:
y = y.cpu()
y = torch.squeeze(y).numpy()
y_label = torch.squeeze(y_label).numpy()
# img_y = y*127.5
if args.choose_net == "Unet":
y = (y > 0.5)
elif args.choose_net == "My_Unet":
y = (y > 0.5)
elif args.choose_net == "Enet":
y = (y > 0.5)
elif args.choose_net == "Segnet":
y = (y > 0.5)
elif args.choose_net == "Scnn":
y = (y > 0.5)
elif args.choose_net == "CascadNet":
y = (y > 0.8)
elif args.choose_net == "my_drsnet_A":
y = (y > 0.5)
elif args.choose_net == "my_drsnet_B":
y = (y > 0.5)
elif args.choose_net == "my_drsnet_C":
y = (y > 0.5)
elif args.choose_net == "my_drsnet_A_direct_skip":
y = (y > 0.5)
elif args.choose_net == "SEResNet":
y = (y > 0.5)
elif args.choose_net == "resnext_unet":
y = (y > 0.5)
elif args.choose_net == "resnet50_unet":
y = (y > 0.5)
elif args.choose_net == "unet_res34":
y = (y > 0.5)
elif args.choose_net == "dfanet":
y = (y > 0.5)
elif args.choose_net == "cgnet":
y = (y > 0.5)
elif args.choose_net == "lednet":
y = (y > 0.5)
elif args.choose_net == "bisenet":
y = (y > 0.5)
elif args.choose_net == "pspnet":
y = (y > 0.5)
elif args.choose_net == "fddwnet":
y = (y > 0.5)
elif args.choose_net == "contextnet":
y = (y > 0.5)
elif args.choose_net == "linknet":
y = (y > 0.5)
elif args.choose_net == "edanet":
y = (y > 0.5)
elif args.choose_net == "erfnet":
y = (y > 0.5)
img_y = y.astype(int).squeeze()
print(y_label.shape, img_y.shape)
image = np.concatenate((img_y, y_label))
y_label = y_label.astype(int)
metric.addBatch(img_y, y_label)
acc = metric.classPixelAccuracy()
mIoU = metric.meanIntersectionOverUnion()
# confusionMatrix=metric.genConfusionMatrix(img_y, y_label)
mean_acc.append(acc[1])
mean_miou.append(mIoU)
# print(acc, mIoU,confusionMatrix)
print(acc, mIoU)
plt.imshow(image * 5)
plt.pause(0.1)
plt.show()
# 计算时需封印acc和miou计算部分
print("Took ", alltime, "seconds")
print("Took", alltime / 638.0, "s/perimage")
print("FPS", 1 / (alltime / 638.0))
print("average acc:%0.6f average miou:%0.6f" %
(np.mean(mean_acc), np.mean(mean_miou)))
def test_img(src_path, label_path):
model_enet = enet.ENet(num_classes=1).to(device)
model_segnet = segnet.SegNet(3, 1).to(device)
model_my_mulSE_A = my_drsnet.MultiscaleSENetA(3, 1).to(device)
model_my_mulSE_B = my_drsnetmy_drsnet.MultiscaleSENetB(3, 1).to(device)
model_my_mulSE_C = my_drsnet.MultiscaleSENetC(3, 1).to(device)
model_my_mulSE_A_direct_skip = my_drsnet.MultiscaleSENetA_direct_skip(
3, 1).to(device)
model_SEResNet18 = my_drsnet.SEResNet18(in_ch=3, out_ch=1).to(device)
ch_cfg = [[8, 48, 96], [240, 144, 288], [240, 144, 288]]
model_dfanet = dfanet.DFANet(ch_cfg, 3, 1).to(device)
model_cgnet = cgnet.Context_Guided_Network(1).to(device)
model_lednet = lednet.Net(num_classes=1).to(device)
model_bisenet = bisenet.BiSeNet(1, 'resnet18').to(device)
model_fddwnet = fddwnet.Net(classes=1).to(device)
model_contextnet = contextnet.ContextNet(classes=1).to(device)
model_linknet = linknet.LinkNet(classes=1).to(device)
model_edanet = edanet.EDANet(classes=1).to(device)
model_erfnet = erfnet.ERFNet(classes=1).to(device)
model_enet.load_state_dict(torch.load("./weight/enet_weight.pth"))
model_enet.eval()
model_segnet.load_state_dict(torch.load("./weight/segnet_weight.pth"))
model_segnet.eval()
model_my_mulSE_A.load_state_dict(
torch.load("./weight/my_drsnet_A_weight.pth"))
model_my_mulSE_A.eval()
model_my_mulSE_B.load_state_dict(
torch.load("./weight/my_drsnet_B_weight.pth"))
model_my_mulSE_B.eval()
model_my_mulSE_C.load_state_dict(
torch.load("./weight/my_drsnet_C_weight.pth"))
model_my_mulSE_C.eval()
model_my_mulSE_A_direct_skip.load_state_dict(
torch.load("./weight/my_drsnet_A_direct_skip_weight.pth"))
model_my_mulSE_A_direct_skip.eval()
model_SEResNet18.load_state_dict(
torch.load("./weight/SEResNet18_weight.pth"))
model_SEResNet18.eval()
model_dfanet.load_state_dict(torch.load("./weight/dfanet.pth"))
model_dfanet.eval()
model_cgnet.load_state_dict(torch.load("./weight/cgnet.pth"))
model_cgnet.eval()
model_lednet.load_state_dict(torch.load("./weight/lednet.pth"))
model_lednet.eval()
model_bisenet.load_state_dict(torch.load("./weight/bisenet.pth"))
model_bisenet.eval()
model_fddwnet.load_state_dict(torch.load("./weight/fddwnet.pth"))
model_fddwnet.eval()
model_contextnet.load_state_dict(torch.load("./weight/contextnet.pth"))
model_contextnet.eval()
model_linknet.load_state_dict(torch.load("./weight/linknet.pth"))
model_linknet.eval()
model_edanet.load_state_dict(torch.load("./weight/edanet.pth"))
model_edanet.eval()
model_erfnet.load_state_dict(torch.load("./weight/erfnet.pth"))
model_erfnet.eval()
src = Image.open(src_path)
src = src.resize((128, 192))
src = x_transform(src)
src = src.to(device)
src = torch.unsqueeze(src, 0)
y_enet = model_enet(src)
# label = label.to(device)
y_enet = y_enet.cpu()
y_enet = y_enet.detach().numpy().reshape(192, 128)
y_segnet = model_segnet(src)
# label = label.to(device)
y_segnet = y_segnet.cpu()
y_segnet = y_segnet.detach().numpy().reshape(192, 128)
y_my_mulSE_A = model_my_mulSE_A(src)
# label = label.to(device)
y_my_mulSE_A = y_my_mulSE_A.cpu()
y_my_mulSE_A = y_my_mulSE_A.detach().numpy().reshape(192, 128)
y_my_mulSE_B = model_my_mulSE_B(src)
# label = label.to(device)
y_my_mulSE_B = y_my_mulSE_B.cpu()
y_my_mulSE_B = y_my_mulSE_B.detach().numpy().reshape(192, 128)
y_my_mulSE_C = model_my_mulSE_C(src)
# label = label.to(device)
y_my_mulSE_C = y_my_mulSE_C.cpu()
y_my_mulSE_C = y_my_mulSE_C.detach().numpy().reshape(192, 128)
y_my_mulSE_A_direct_skip = model_my_mulSE_A_direct_skip(src)
# label = label.to(device)
y_my_mulSE_A_direct_skip = y_my_mulSE_A_direct_skip.cpu()
y_my_mulSE_A_direct_skip = y_my_mulSE_A_direct_skip.detach().numpy(
).reshape(192, 128)
y_SEResNet18 = model_SEResNet18(src)
# label = label.to(device)
y_SEResNet18 = y_SEResNet18.cpu()
y_SEResNet18 = y_SEResNet18.detach().numpy().reshape(192, 128)
y_dfanet = model_dfanet(src)
# label = label.to(device)
y_dfanet = y_dfanet.cpu()
y_dfanet = y_dfanet.detach().numpy().reshape(192, 128)
y_cgnet = model_cgnet(src)
# label = label.to(device)
y_cgnet = y_cgnet.cpu()
y_cgnet = y_cgnet.detach().numpy().reshape(192, 128)
y_lednet = model_lednet(src)
# label = label.to(device)
y_lednet = y_lednet.cpu()
y_lednet = y_lednet.detach().numpy().reshape(192, 128)
y_bisenet = model_bisenet(src)
# label = label.to(device)
y_bisenet = y_bisenet.cpu()
y_bisenet = y_bisenet.detach().numpy().reshape(192, 128)
y_fddwnet = model_fddwnet(src)
# label = label.to(device)
y_fddwnet = y_fddwnet.cpu()
y_fddwnet = y_fddwnet.detach().numpy().reshape(192, 128)
y_contextnet = model_contextnet(src)
# label = label.to(device)
y_contextnet = y_contextnet.cpu()
y_contextnet = y_contextnet.detach().numpy().reshape(192, 128)
y_linknet = model_linknet(src)
# label = label.to(device)
y_linknet = y_linknet.cpu()
y_linknet = y_linknet.detach().numpy().reshape(192, 128)
y_edanet = model_edanet(src)
# label = label.to(device)
y_edanet = y_edanet.cpu()
y_edanet = y_edanet.detach().numpy().reshape(192, 128)
y_erfnet = model_erfnet(src)
# label = label.to(device)
y_erfnet = y_erfnet.cpu()
y_erfnet = y_erfnet.detach().numpy().reshape(192, 128)
y_enet = (y_enet > 0.5).astype(int) * 255
y_segnet = (y_segnet > 0.5).astype(int) * 255
y_my_mulSE_A = (y_my_mulSE_A > 0.5).astype(int) * 255
y_my_mulSE_B = (y_my_mulSE_B > 0.5).astype(int) * 255
y_my_mulSE_C = (y_my_mulSE_C > 0.5).astype(int) * 255
y_my_mulSE_A_direct_skip = (y_my_mulSE_A_direct_skip >
0.5).astype(int) * 255
y_SEResNet18 = (y_SEResNet18 > 0.5).astype(int) * 255
y_dfanet = (y_dfanet > 0.5).astype(int) * 255
y_cgnet = (y_cgnet > 0.5).astype(int) * 255
y_lednet = (y_lednet > 0.5).astype(int) * 255
y_bisenet = (y_bisenet > 0.5).astype(int) * 255
y_fddwnet = (y_fddwnet > 0.5).astype(int) * 255
y_contextnet = (y_contextnet > 0.5).astype(int) * 255
y_linknet = (y_linknet > 0.5).astype(int) * 255
y_edanet = (y_edanet > 0.5).astype(int) * 255
y_erfnet = (y_erfnet > 0.5).astype(int) * 255
src1 = Image.open(src_path)
src1 = src1.resize((128, 192))
label = Image.open(label_path)
label = label.resize((128, 192))
label = np.array(label) * 255
src1.save("./data/result/" + "_src.png")
cv2.imwrite("./data/result/" + "_label.png", label)
cv2.imwrite("./data/result/" + "enet_predict.png", y_enet)
cv2.imwrite("./data/result/" + "segnet_predict.png", y_segnet)
cv2.imwrite("./data/result/" + "my_drsnet_A_predict.png", y_my_mulSE_A)
cv2.imwrite("./data/result/" + "my_drsnet_B_predict.png", y_my_mulSE_B)
cv2.imwrite("./data/result/" + "my_drsnet_C_predict.png", y_my_mulSE_C)
cv2.imwrite("./data/result/" + "my_drsnet_A_direct_skip_predict.png",
y_my_mulSE_A_direct_skip)
cv2.imwrite("./data/result/" + "y_SEResNet18_predict.png", y_SEResNet18)
cv2.imwrite("./data/result/" + "dfanet_predict.png", y_dfanet)
cv2.imwrite("./data/result/" + "cgnet_predict.png", y_cgnet)
cv2.imwrite("./data/result/" + "lednet_predict.png", y_lednet)
cv2.imwrite("./data/result/" + "bisenet_predict.png", y_bisenet)
cv2.imwrite("./data/result/" + "fddwnet_predict.png", y_fddwnet)
cv2.imwrite("./data/result/" + "contextnet_predict.png", y_contextnet)
cv2.imwrite("./data/result/" + "linknet_predict.png", y_linknet)
cv2.imwrite("./data/result/" + "edanet_predict.png", y_edanet)
cv2.imwrite("./data/result/" + "erfnet_predict.png", y_erfnet)
return 0
def train():
if args.choose_net == "Unet":
model = my_unet.UNet(3, 1).to(device)
if args.choose_net == "My_Unet":
model = my_unet.My_Unet2(3, 1).to(device)
elif args.choose_net == "Enet":
model = enet.ENet(num_classes=13).to(device)
elif args.choose_net == "Segnet":
model = segnet.SegNet(3, 13).to(device)
elif args.choose_net == "CascadNet":
model = my_cascadenet.CascadeNet(3, 1).to(device)
elif args.choose_net == "my_drsnet_A":
model = my_drsnet.MultiscaleSENetA(in_ch=3, out_ch=1).to(device)
elif args.choose_net == "my_drsnet_B":
model = my_drsnet.MultiscaleSENetB(in_ch=3, out_ch=1).to(device)
elif args.choose_net == "my_drsnet_C":
model = my_drsnet.MultiscaleSENetC(in_ch=3, out_ch=1).to(device)
elif args.choose_net == "my_drsnet_A_direct_skip":
model = my_drsnet.MultiscaleSENetA_direct_skip(in_ch=3,
out_ch=1).to(device)
elif args.choose_net == "SEResNet":
model = my_drsnet.SEResNet18(in_ch=3, out_ch=1).to(device)
elif args.choose_net == "resnext_unet":
model = resnext_unet.resnext50(in_ch=3, out_ch=1).to(device)
elif args.choose_net == "resnet50_unet":
model = resnet50_unet.UNetWithResnet50Encoder(in_ch=3,
out_ch=1).to(device)
elif args.choose_net == "unet_nest":
model = unet_nest.UNet_Nested(3, 2).to(device)
elif args.choose_net == "unet_res34":
model = unet_res34.Resnet_Unet(3, 1).to(device)
elif args.choose_net == "trangle_net":
model = mytrangle_net.trangle_net(3, 1).to(device)
elif args.choose_net == "dfanet":
ch_cfg = [[8, 48, 96], [240, 144, 288], [240, 144, 288]]
model = dfanet.DFANet(ch_cfg, 3, 1).to(device)
elif args.choose_net == "lednet":
model = lednet.Net(num_classes=1).to(device)
elif args.choose_net == "cgnet":
model = cgnet.Context_Guided_Network(classes=1).to(device)
elif args.choose_net == "pspnet":
model = pspnet.PSPNet(1).to(device)
elif args.choose_net == "bisenet":
model = bisenet.BiSeNet(1, 'resnet18').to(device)
elif args.choose_net == "espnet":
model = espnet.ESPNet(classes=1).to(device)
elif args.choose_net == "fddwnet":
model = fddwnet.Net(classes=1).to(device)
elif args.choose_net == "contextnet":
model = contextnet.ContextNet(classes=1).to(device)
elif args.choose_net == "linknet":
model = linknet.LinkNet(classes=1).to(device)
elif args.choose_net == "edanet":
model = edanet.EDANet(classes=1).to(device)
elif args.choose_net == "erfnet":
model = erfnet.ERFNet(classes=1).to(device)
from collections import OrderedDict
loadpretrained = 0
# 0:no loadpretrained model
# 1:loadpretrained model to original network
# 2:loadpretrained model to new network
if loadpretrained == 1:
model.load_state_dict(torch.load(args.weight))
elif loadpretrained == 2:
model = my_drsnet.MultiscaleSENetA(in_ch=3, out_ch=1).to(device)
model_dict = model.state_dict()
pretrained_dict = torch.load(args.weight)
pretrained_dict = {
k: v
for k, v in pretrained_dict.items() if k in model_dict
}
model_dict.update(pretrained_dict)
model.load_state_dict(model_dict)
# model.load_state_dict(torch.load(args.weight))
# pretrained_dict = {k: v for k, v in model.items() if k in model} # filter out unnecessary keys
# model.update(pretrained_dict)
# model.load_state_dict(model)
# 计算模型参数量和计算量FLOPs
dsize = (1, 3, 128, 192)
inputs = torch.randn(dsize).to(device)
total_ops, total_params = profile(model, (inputs, ), verbose=False)
print(" %.2f | %.2f" % (total_params / (1000**2), total_ops / (1000**3)))
batch_size = args.batch_size
# 加载数据集
liver_dataset = LiverDataset("data/train_camvid/",
transform=x_transform,
target_transform=y_transform)
len_img = liver_dataset.__len__()
dataloader = DataLoader(liver_dataset,
batch_size=batch_size,
shuffle=True,
num_workers=24)
# DataLoader:该接口主要用来将自定义的数据读取接口的输出或者PyTorch已有的数据读取接口的输入按照batch size封装成Tensor
# batch_size：how many samples per minibatch to load，这里为4，数据集大小400，所以一共有100个minibatch
# shuffle:每个epoch将数据打乱，这里epoch=10。一般在训练数据中会采用
# num_workers：表示通过多个进程来导入数据，可以加快数据导入速度
# 梯度下降
# optimizer = optim.Adam(model.parameters()) # model.parameters():Returns an iterator over module parameters
# # Observe that all parameters are being optimized
optimizer = optim.SGD(model.parameters(),
lr=args.lr,
momentum=0.9,
weight_decay=0.0001)
# 每n个epoches来一次余弦退火
cosine_lr_scheduler = lr_scheduler.CosineAnnealingLR(
optimizer, T_max=10 * int(len_img / batch_size), eta_min=0.00001)
multiclass = 1
if multiclass == 1:
# 损失函数
class_weights = np.array([
0., 6.3005947, 4.31063664, 34.09234699, 50.49834979, 3.88280945,
50.49834979, 8.91626081, 47.58477105, 29.41289083, 18.95706775,
37.84558871, 39.3477858
]) #camvid
# class_weights = weighing(dataloader, 13, c=1.02)
class_weights = torch.from_numpy(class_weights).float().to(device)
criterion = torch.nn.CrossEntropyLoss(weight=class_weights)
# criterion = LovaszLossSoftmax()
# criterion = torch.nn.MSELoss()
train_modelmulticlasses(model, criterion, optimizer, dataloader,
cosine_lr_scheduler)
else:
# 损失函数
# criterion = LovaszLossHinge()
# weights=[0.2]
# weights=torch.Tensor(weights).to(device)
# # criterion = torch.nn.CrossEntropyLoss(weight=weights)
criterion = torch.nn.BCELoss()
# criterion =focal_loss.FocalLoss(1)
train_model(model, criterion, optimizer, dataloader,
cosine_lr_scheduler)
n_classes = args.nClasses
images_path = '../../datasets/segmentation/'
val_file = './data/seg_test.txt' if n_classes == 2 else './data/parse_test.txt'
weights_file = './weights/{}_seg_weights.h5'.format(args.model) if n_classes == 2 \
else './weights/{}_parse_weights.h5'.format(args.model)
input_height = 256
input_width = 256
if args.model == 'unet':
m = unet.Unet(n_classes,
input_height=input_height,
input_width=input_width)
elif args.model == 'segnet':
m = segnet.SegNet(n_classes,
input_height=input_height,
input_width=input_width)
else:
raise ValueError('Do not support {}'.format(args.model))
m.load_weights(weights_file.format(args.model))
m.compile(loss='categorical_crossentropy',
optimizer='adam',
metrics=['accuracy'])
print('Start evaluating..')
pbdr = tqdm(total=5000)
iou = [0. for _ in range(1, n_classes)]
count = [0. for _ in range(1, n_classes)]
for x, y in generator(images_path,
val_file,
def __init__(self, args):
self.model = segnet.SegNet(
args.in_ch, args.out_ch, args.base_kernel)
def test_backward(self):
opt = optimizers.MomentumSGD(lr=0.01)
# Update each depth
for depth in six.moves.range(1, self.n_encdec + 1):
model = segnet.SegNet(self.n_encdec, self.n_classes,
self.x_shape[1], self.n_mid)
model = segnet.SegNetLoss(model,
class_weight=None,
train_depth=depth)
opt.setup(model)
# Deregister non-target links from opt
if depth > 1:
model.predictor.remove_link('conv_cls')
for d in range(1, self.n_encdec + 1):
if d != depth:
model.predictor.remove_link('encdec{}'.format(d))
# Keep the initial values
prev_params = {
'conv_cls': copy.deepcopy(model.predictor.conv_cls.W.data)
}
for d in range(1, self.n_encdec + 1):
name = '/encdec{}/enc/W'.format(d)
encdec = getattr(model.predictor, 'encdec{}'.format(d))
prev_params[name] = copy.deepcopy(encdec.enc.W.data)
self.assertTrue(prev_params[name] is not encdec.enc.W.data)
# Update the params
x, t = self.get_xt()
loss = model(x, t)
loss.data *= 1E20
model.cleargrads()
loss.backward()
opt.update()
for d in range(1, self.n_encdec + 1):
# The weight only in the target layer should be updated
c = self.assertFalse if d == depth else self.assertTrue
encdec = getattr(opt.target.predictor, 'encdec{}'.format(d))
self.assertTrue(hasattr(encdec, 'enc'))
self.assertTrue(hasattr(encdec.enc, 'W'))
self.assertTrue('/encdec{}/enc/W'.format(d) in prev_params)
c(np.array_equal(encdec.enc.W.data,
prev_params['/encdec{}/enc/W'.format(d)]),
msg='depth:{} d:{} diff:{}'.format(
depth, d,
np.sum(encdec.enc.W.data -
prev_params['/encdec{}/enc/W'.format(d)])))
if depth == 1:
# The weight in the last layer should be updated
self.assertFalse(
np.allclose(model.predictor.conv_cls.W.data,
prev_params['conv_cls']))
cg = build_computational_graph([loss],
variable_style=_var_style,
function_style=_func_style).dump()
fn = 'tests/SegNet_bw_depth-{}_{}.dot'.format(self.n_encdec, depth)
if os.path.exists(fn):
continue
with open(fn, 'w') as f:
f.write(cg)
subprocess.call('dot -Tpng {} -o {}'.format(
fn, fn.replace('.dot', '.png')),
shell=True)
for name, param in model.namedparams():
encdec_depth = re.search('encdec([0-9]+)', name)
if encdec_depth:
ed = int(encdec_depth.groups()[0])
self.assertEqual(ed, depth)
