fastreid.modeling.backbones.osnet.OSNet Class Reference

Inheritance diagram for fastreid.modeling.backbones.osnet.OSNet:

Public Member Functions
	__init__ (self, blocks, layers, channels, bn_norm, IN=False, **kwargs)
	forward (self, x)

Public Attributes
	conv1
	maxpool
	conv2
	conv3
	conv4
	conv5

Protected Member Functions
	_make_layer (self, block, layer, in_channels, out_channels, bn_norm, reduce_spatial_size, IN=False)
	_init_params (self)

Detailed Description

Omni-Scale Network.

Reference:
    - Zhou et al. Omni-Scale Feature Learning for Person Re-Identification. ICCV, 2019.
    - Zhou et al. Learning Generalisable Omni-Scale Representations
      for Person Re-Identification. arXiv preprint, 2019.

Definition at line 289 of file osnet.py.

Constructor & Destructor Documentation

__init__()

fastreid.modeling.backbones.osnet.OSNet.__init__	(		self,
			blocks,
			layers,
			channels,
			bn_norm,
			IN = False,
		**	kwargs )

Definition at line 298 of file osnet.py.

    ):
        super(OSNet, self).__init__()
        num_blocks = len(blocks)
        assert num_blocks == len(layers)
        assert num_blocks == len(channels) - 1
 
        # convolutional backbone
        self.conv1 = ConvLayer(3, channels[0], 7, bn_norm, stride=2, padding=3, IN=IN)
        self.maxpool = nn.MaxPool2d(3, stride=2, padding=1)
        self.conv2 = self._make_layer(
            blocks[0],
            layers[0],
            channels[0],
            channels[1],
            bn_norm,
            reduce_spatial_size=True,
            IN=IN
        )
        self.conv3 = self._make_layer(
            blocks[1],
            layers[1],
            channels[1],
            channels[2],
            bn_norm,
            reduce_spatial_size=True
        )
        self.conv4 = self._make_layer(
            blocks[2],
            layers[2],
            channels[2],
            channels[3],
            bn_norm,
            reduce_spatial_size=False
        )
        self.conv5 = Conv1x1(channels[3], channels[3], bn_norm)
 
        self._init_params()
 

Member Function Documentation

_init_params()

fastreid.modeling.backbones.osnet.OSNet._init_params ( self )

protected

Definition at line 370 of file osnet.py.

    def _init_params(self):
        for m in self.modules():
            if isinstance(m, nn.Conv2d):
                nn.init.kaiming_normal_(
                    m.weight, mode='fan_out', nonlinearity='relu'
                )
                if m.bias is not None:
                    nn.init.constant_(m.bias, 0)
 
            elif isinstance(m, nn.BatchNorm2d):
                nn.init.constant_(m.weight, 1)
                nn.init.constant_(m.bias, 0)
 
            elif isinstance(m, nn.BatchNorm1d):
                nn.init.constant_(m.weight, 1)
                nn.init.constant_(m.bias, 0)
 
            elif isinstance(m, nn.Linear):
                nn.init.normal_(m.weight, 0, 0.01)
                if m.bias is not None:
                    nn.init.constant_(m.bias, 0)
 

_make_layer()

fastreid.modeling.backbones.osnet.OSNet._make_layer ( self, block, layer, in_channels, out_channels, bn_norm, reduce_spatial_size, IN = False )

protected

Definition at line 344 of file osnet.py.

    ):
        layers = []
 
        layers.append(block(in_channels, out_channels, bn_norm, IN=IN))
        for i in range(1, layer):
            layers.append(block(out_channels, out_channels, bn_norm, IN=IN))
 
        if reduce_spatial_size:
            layers.append(
                nn.Sequential(
                    Conv1x1(out_channels, out_channels, bn_norm),
                    nn.AvgPool2d(2, stride=2),
                )
            )
 
        return nn.Sequential(*layers)
 

forward()

fastreid.modeling.backbones.osnet.OSNet.forward	(		self,
			x )

Definition at line 392 of file osnet.py.

    def forward(self, x):
        x = self.conv1(x)
        x = self.maxpool(x)
        x = self.conv2(x)
        x = self.conv3(x)
        x = self.conv4(x)
        x = self.conv5(x)
        return x
 
 

Member Data Documentation

conv1

fastreid.modeling.backbones.osnet.OSNet.conv1

Definition at line 313 of file osnet.py.

conv2

fastreid.modeling.backbones.osnet.OSNet.conv2

Definition at line 315 of file osnet.py.

conv3

fastreid.modeling.backbones.osnet.OSNet.conv3

Definition at line 324 of file osnet.py.

conv4

fastreid.modeling.backbones.osnet.OSNet.conv4

Definition at line 332 of file osnet.py.

conv5

fastreid.modeling.backbones.osnet.OSNet.conv5

Definition at line 340 of file osnet.py.

maxpool

fastreid.modeling.backbones.osnet.OSNet.maxpool

Definition at line 314 of file osnet.py.

---

The documentation for this class was generated from the following file:

• smreid/fastreid/modeling/backbones/osnet.py