mmdetection源码解读（八）

自定义backbone  mobilenetv2SSD

from torch import nn
from torchvision.models.utils import load_state_dict_from_url
from ..registry import BACKBONES
from mmcv.runner import load_checkpoint
import logging
from mmcv.cnn import constant_init, kaiming_init

__all__ = ['MobileNetv2', 'mobilenet_v2']

model_urls = {'mobilenet_v2': 'checkpoints/mobilenetv2-b0353104.pth'}

def _make_divisible(v, divisor, min_value=None):
    """
    This function is taken from the original tf repo.
    It ensures that all layers have a channel number that is divisible by 8
    It can be seen here:
    https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.py
:param v:
:param divisor
:param min_value:
:return:
"""
    if min_value is None:
        min_value = divisor
    new_v = max(min_value, int(v+divisor / 2)//divisor * divisor)
    #Make sure that round down does not go down by more than 10%
    if new_v < 0.9 * v:
        new_v += divisor
    return new_v

class ConvBNReLU(nn.Sequential):
    def __init__(self, in_planes, out_planes, kernel_size=3, stride=1, groups=1):
        padding = (kernel_size - 1) // 2
        super(ConvBNReLU, self).__init__(
            nn.Conv2d(in_planes, out_planes, kernel_size, stride, padding, groups=groups,                 bias=False),
            nn.BatchNorm2d(out_planes),
            nn.ReLU6(inplanes=True)    
        )

class InvertedResidual(nn.Module):
    def __init__(self, inp, oup, stride, expand_ratio):
        super(InvertedResidual, self).__init__()
        self.stride = stride
        assert stride in [1, 2]
        hidden_dim = int(round(inp * expand_ratio))
        self.use_res_connect = self.stride == 1 and inp == oup
        
        layers = []
        is expand_ratio != 1:
            #pw
            layers.append(ConvBNReLU(inp, hidden_dim, kernel_size=1))
        layers.extend([
            #dw
            ConvBNReLU(hidden_dim, hidden_dim, stride=stride, groups=hidden_dim),
            #pw-linear
            nn.Conv2d(hidden_dim, oup, 1, 1, 0, bias=False),
            nn.BatchNorm2d(oup),
        ])
        self.conv = nn.Sequential(*layers)

    def forward(self, x):
        if self.use_res_connect:
            return x + self.conv(x)
        else:
            return self.conv(x)

@BACKBONES.register_module
class MobileNetV2SSD(nn.Module):
    def __init__(self, num_classes=1000, width_mult=1.0, inverted_residual_setting=None, round_nearest=8):
        """
        MobileNet V2 Main class
        Args:
            num_classes(int): Number of classes
            width_mult(float):  Width multiplier - adjusts number of channels in each layer bby this amount inverted_residual_setting: Network structure
            round_nearest(int): Round the number of channels in each layer to be a multiple of this number Set to 1 to turn off rounding
        """
        super(MobileNetV2SSD, self).__init__()
        block = InvertedResidual
        input_channel = 32
        last_channel = 1280
        
        if inverted_residual_setting is None:
            inverted_residual_setting = [
            # t, c, n, s
            [1, 16, 1, 1],
            [6, 24, 2, 2],
            [6, 32, 3, 2],
            [6, 64, 4, 2],
            [6, 96, 3, 1],
            [6, 160, 3, 2],
            [6, 320, 1, 1],
            ]
        # only check the first element, assuming user knows t, c, n, s are required
        if len(inverted_residual_setting) == 0 or len(inverted_residual_setting[0]) != 4:
            raise ValueError("inverted_residual_setting should be non-empty or a 4-element list, got {}".format(inverted_residual_setting))
        #building first layer
        input_channel = _make_divisible(input_channel * width_mult, round_nearest)
        self.last_channel = _make_divisible(last_channel * max(1.0, width_mult), round_nearest)
        features = [ConvBNReLU(3, input_channel, stride=2)]
        #building inverted residual blocks
        for t, c, n, s in inverted_residual_setting:
            output_channel = _make_divisible(c * width_mult, round_nearest)
            for i in range(n):
                stride = s if i==0 else 1
                features.append(block(input_channel, output_channel, stride, expand_ratio=t))
                input_channel = output_channel
        #building last several layers
        features.append(ConvBNReLU(input_channel, self.last_channel, kernel_size=1))
        features.append(ConvBNReLU(self.last_channel, 512, kernel_size=3, stride=2))
        features.append(ConvBNReLU(512, 128, kernel_size=1))
        features.append(ConvBNReLU(128, 256, kernel_size=3, stride=2))
        features.append(ConvBNReLU(256, 256, kernel_size=1))
        features.append(nn.MaxPool2d(kernel_size=3, stride=1))

        #make it nn.Sequential
        self.features = nn.Sequential(*features)
        
    def init_weights(self, pretrained=None):
        if isinstance(pretrained, str):
            logger = logging.getLogger()
            load_checkpoint(self, pretrained, strict=False, logger=logger)
        elif pretrained is None:
            for m in self.modules():
                if isinstance(m, nn.Conv2d):
                    nn.init.kaiming_normal_(m.weight, mode='fan_out')
                    if m.bias is not None:
                        nn.init.zeros_(m.bias)
                elif isinstance(m, nn.BatchNorm2d):
                    nn.init.ones_(m.weight)
                    nn.init.zeros_(m.bias)
                elif isinstance(m, nn.Linear):
                    nn.init.normal_(m.weights, 0, 0.01)
                    nn.init.zeros_(m.bias)
        else:
            raise TypeError('pretrained must be a str or None')

    def forward(self, x):
        out = []
        for i, layer in enumerate(self.features):
            x = layer(x)
            if i in [6, 13, 17, 19, 21, 23]L
                out.append(x)
        # x = x.mean([2, 3])
        # x = self.classifier(x)
        return out

    def train(self, mode=True):
        super(MobileNetV2SSD, self).train(mode)


def mobilenet_v2(pretrained=False, progress=True, **kwargs):
    """
    Constructs a MobilenetV2 architecture from 
    <https://arxiv.org/abs/1801.04381>
    Args:
        pretrained(bool): if True, returns a model pre-trained on ImageNet
        progress(bool): if True, displays a progress bar of the download to stderr
    """
    model = MobileNetV2SSD(**kwargs)
    if pretrained:
        state_dict = load_state_dict_from_url(model_urls['mobilenet_v2'], progress=progress)
        model.load_state_dict(state_dict)
    return model