brails.processors.nfloors_detector.lib.efficientdet.model module

class brails.processors.nfloors_detector.lib.efficientdet.model.BiFPN(num_channels, conv_channels, first_time=False, epsilon=0.0001, onnx_export=False, attention=True)

Bases: Module

modified by Zylo117

forward(inputs)
illustration of a minimal bifpn unit
P7_0 ————————-> P7_2 ——–>
|-------------|

↓ |

P6_0 ———> P6_1 ———> P6_2 ——–>
|-------------|————–↑ ↑

↓ |

P5_0 ———> P5_1 ———> P5_2 ——–>
|-------------|————–↑ ↑

↓ |

P4_0 ———> P4_1 ———> P4_2 ——–>
|-------------|————–↑ ↑

|————–↓ |

P3_0 ————————-> P3_2 ——–>

class brails.processors.nfloors_detector.lib.efficientdet.model.Classifier(in_channels, num_anchors, num_classes, num_layers, onnx_export=False)

Bases: Module

modified by Zylo117

forward(inputs)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

class brails.processors.nfloors_detector.lib.efficientdet.model.EfficientNet(compound_coef, load_weights=False)

Bases: Module

modified by Zylo117

forward(x)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

class brails.processors.nfloors_detector.lib.efficientdet.model.Regressor(in_channels, num_anchors, num_layers, onnx_export=False)

Bases: Module

modified by Zylo117

forward(inputs)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

class brails.processors.nfloors_detector.lib.efficientdet.model.SeparableConvBlock(in_channels, out_channels=None, norm=True, activation=False, onnx_export=False)

Bases: Module

created by Zylo117

forward(x)

Define the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

brails.processors.nfloors_detector.lib.efficientdet.model.nms(dets, thresh)