def create_cnn_model(base_arch,
nc: int,
cut: Union[int, Callable] = None,
pretrained: bool = True,
lin_ftrs: Optional[Collection[int]] = None,
ps: Floats = 0.5,
custom_head: Optional[nn.Module] = None,
bn_final: bool = False,
concat_pool: bool = True):
"Create custom convnet architecture"
print(type(base_arch), base_arch)
body = create_body(base_arch, pretrained, cut)
# body.requires_grad = True
nf = num_features_model(
nn.Sequential(*body.children())) * (2 if concat_pool else 1)
head = create_head(nf,
nc,
ps=ps,
concat_pool=concat_pool,
bn_final=bn_final)
model = nn.Sequential(body, head)
return model
def create_cnn(data, arch, pretrained=False, is_mono_input=True, **kwargs):
meta = cnn_config(arch)
body = create_body(arch, pretrained)
# sum up the weights of in_channels axis, to reduce to single input channel
# Suggestion by David Gutman
# https://forums.fast.ai/t/black-and-white-images-on-vgg16/2479/2
if is_mono_input:
first_conv_layer = body[0]
first_conv_weights = first_conv_layer.state_dict()['weight']
assert first_conv_weights.size(1) == 3 # RGB channels dim
summed_weights = torch.sum(first_conv_weights, dim=1, keepdim=True)
first_conv_layer.weight.data = summed_weights
first_conv_layer.in_channels = 1
else:
# In this case, the input is a stereo
first_conv_layer = body[0]
first_conv_weights = first_conv_layer.state_dict()['weight']
assert first_conv_weights.size(1) == 3 # RGB channels dim
summed_weights = torch.sum(first_conv_weights, dim=1, keepdim=True)
first_conv_layer.weight.data = first_conv_weights[:, :
2, :, :] # Keep only 2 channels for the weights
first_conv_layer.in_channels = 2
nf = num_features_model(body) * 2
head = create_head(nf, data.c, None, 0.5)
model = nn.Sequential(body, head)
learn = Learner(data, model, **kwargs)
learn.split(meta['split'])
if pretrained:
learn.freeze()
apply_init(model[1], nn.init.kaiming_normal_)
return learn
def __init__(self,
base_arch,
no_diseases,
dropout=0.5,
init=nn.init.kaiming_normal_):
super(CNNPretrainedModel, self).__init__()
self.body = create_body(base_arch, pretrained=True)
nf = num_features_model(nn.Sequential(*self.body.children())) * 2
self.disease_head = create_head(nf,
no_diseases,
ps=0.5,
concat_pool=True,
bn_final=False)
#self.age_head = create_head(nf, 1, ps=0.5, concat_pool=True, bn_final=False)
self.gender_head = create_head(nf,
2,
ps=0.5,
concat_pool=True,
bn_final=False)
#self.projection_head = create_head(nf, 3, ps=0.5, concat_pool=True, bn_final=False)
self.disease_model = nn.Sequential(self.body, self.disease_head)
self.meta = cnn_config(base_arch)
self.split(self.meta['split'])
self.freeze()
apply_init(self.disease_head, init)
#apply_init(self.age_head, init)
apply_init(self.gender_head, init)
def __init__(self, config):
super().__init__()
self.n_emb = config.model.n_emb
self.radius = config.model.radius
self.n_class = config.model.n_class
self.body = get_body(config)
nf = num_features_model(nn.Sequential(*self.body.children())) * 2
self.head = create_head(nf, self.n_emb, lin_ftrs=[1024], ps=config.model.drop_rate, concat_pool=True, bn_final=True)
self.cos_sim = CosSimCenters(self.n_emb, self.n_class)
def __init__(self,
encoder: nn.Module,
n_classes,
final_bias: float = 0.,
n_conv: float = 4,
chs=256,
n_anchors=9,
flatten=True,
sizes=None):
super().__init__(encoder, n_classes, final_bias, n_conv, chs,
n_anchors, flatten, sizes)
self.classifier_image = self._create_image_classifier(
nf=num_features_model(self.encoder) * 2,
nc=1,
y_range=[0 - 0.5, n_classes - 0.5])
self.box_reg_classifier = self._head_box_reg_subnet(
n_classes=1,
n_anchors=n_anchors,
n_conv=n_conv,
chs=chs,
y_range=[0 - 0.5, n_classes - 0.5])
def create_cnn_2(arch: Callable,
n_classes: int = 2,
cut: Union[int, Callable] = None,
pretrained: bool = True,
lin_ftrs: Optional[Collection[int]] = None,
ps: Floats = 0.5,
custom_head: Optional[nn.Module] = None,
split_on: Optional[SplitFuncOrIdxList] = None,
classification: bool = True,
**kwargs: Any) -> Learner:
"Build convnet style learners."
assert classification, 'Regression CNN not implemented yet, bug us on the forums if you want this!'
meta = {'cut': -2, 'split': _resnet_split_2}
body = create_body(arch(pretrained), ifnone(cut, meta['cut']))
nf = num_features_model(body) * 2
head = custom_head or create_head(nf, n_classes, lin_ftrs, ps)
model = nn.Sequential(body, head)
#learner_cls = ifnone(data.learner_type(), ClassificationLearner_2)
learner_cls = ClassificationLearner_2
learn = learner_cls(data, model, **kwargs)
learn.split(ifnone(split_on, meta['split']))
if pretrained: learn.freeze()
apply_init(model[1], nn.init.kaiming_normal_)
return learn
def __init__(self, arch=models.resnet18):
super().__init__()
self.cnn = create_body(arch)
self.head = create_head(num_features_model(self.cnn) * 2, 4)
def resnet(classes):
base_model = models.resnet50(pretrained=True)
body = nn.Sequential(*list(base_model.children())[:-2])
nf = num_features_model(body) * 2
head = create_head(nf, classes, None, ps=0.5, bn_final=False)
return nn.Sequential(body, head, nn.LogSoftmax())
