How can I train the FoveaBox with mobilenet ?

[xscjun]

I add a mobilenet.py as GETTING_STARTED.md said , and modify the config ,but I can't run the training, how can I train the foveabox with mobilenet,please ?

[taokong]

I add a mobilenet.py as GETTING_STARTED.md said , and modify the config ,but I can't run the training, how can I train the foveabox with mobilenet,please ?

Please provide more information?

[xscjun]

I add a mobilenet.py as GETTING_STARTED.md said , and modify the config ,but I can't run the training, how can I train the foveabox with mobilenet,please ?

Please provide more information?

1、my config :fovea_align_gn_ms_mobilenet_fpn_4gpu_2x.py model = dict( type='FOVEA', pretrained=None, backbone=dict( type='MobileNet' # width_multiplier=1, # class_num=1 ), neck=dict( type='FPN', # in_channels=[256, 512, 1024, 2048], # mobilenetweb in_channels=[128, 256, 512, 1024], out_channels=256, start_level=1, num_outs=5, add_extra_convs=True), bbox_head=dict( type='FoveaHead', ... )
nothing else changed

2、backbone : Mobilenet import torch import torch.nn as nn import logging import torch.utils.checkpoint as cp

from mmcv.cnn import constant_init, kaiming_init from mmcv.runner import load_checkpoint from ..registry import BACKBONES

class DepthSeperabelConv2d(nn.Module): def init(self, input_channels, output_channels, kernel_size, **kwargs): super().init() self.depthwise = nn.Sequential( nn.Conv2d( input_channels, input_channels, kernel_size, groups=input_channels, **kwargs), nn.BatchNorm2d(input_channels), nn.ReLU(inplace=True) )

    self.pointwise = nn.Sequential(
        nn.Conv2d(input_channels, output_channels, 1),
        nn.BatchNorm2d(output_channels),
        nn.ReLU(inplace=True)
    )

def forward(self, x):
    x = self.depthwise(x)
    x = self.pointwise(x)
    return x

class BasicConv2d(nn.Module): def init(self, input_channels, output_channels, kernel_size, **kwargs):

    super().init()
    self.conv = nn.Conv2d(input_channels, output_channels, kernel_size, **kwargs)
    self.bn   = nn.BatchNorm2d(output_channels)
    self.relu = nn.ReLU(inplace=True)

    def forward(self, x):
        x = self.conv(x)
        x = self.bn(x)
        x = self.relu(x)

    return x

@BACKBONES.register_module class MobileNet(nn.Module): def init(self, width_multiplier=1, class_num=1): super(MobileNet, self).init() alpha = width_multiplier self.stem = nn.Sequential( BasicConv2d(3, int(32 * alpha), 3, padding=1, bias=False), DepthSeperabelConv2d( int(32 * alpha), int(64 * alpha), 3, padding=1, bias=False ) )

    #downsample
    self.conv1 = nn.Sequential(
        DepthSeperabelConv2d(
            int(64 * alpha),
            int(128 * alpha),
            3,
            stride=2,
            padding=1,
            bias=False
        ),
        DepthSeperabelConv2d(
            int(128 * alpha),
            int(128 * alpha),
            3,
            padding=1,
            bias=False
        )
    )
    #downsample
    self.conv2 = nn.Sequential(
        DepthSeperabelConv2d(
            int(128 * alpha),
            int(256 * alpha),
            3,
            stride=2,
            padding=1,
            bias=False
        ),
        DepthSeperabelConv2d(
            int(256 * alpha),
            int(256 * alpha),
            3,
            padding=1,
            bias=False
        )
    )
    #downsample
    self.conv3 = nn.Sequential(
        DepthSeperabelConv2d(
            int(256 * alpha),
            int(512 * alpha),
            3,
            stride=2,
            padding=1,
            bias=False
        ),
        DepthSeperabelConv2d(
            int(512 * alpha),
            int(512 * alpha),
            3,
            padding=1,
            bias=False
        ),
        DepthSeperabelConv2d(
            int(512 * alpha),
            int(512 * alpha),
            3,
            padding=1,
            bias=False
        ),
        DepthSeperabelConv2d(
            int(512 * alpha),
            int(512 * alpha),
            3,
            padding=1,
            bias=False
        ),
        DepthSeperabelConv2d(
            int(512 * alpha),
            int(512 * alpha),
            3,
            padding=1,
            bias=False
        ),
        DepthSeperabelConv2d(
            int(512 * alpha),
            int(512 * alpha),
            3,
            padding=1,
            bias=False
        )
    )
    #downsample
    self.conv4 = nn.Sequential(
        DepthSeperabelConv2d(
            int(512 * alpha),
            int(1024 * alpha),
            3,
            stride=2,
            padding=1,
            bias=False
        ),
        DepthSeperabelConv2d(
            int(1024 * alpha),
            int(1024 * alpha),
            3,
            padding=1,
            bias=False
        )
    )
    # self.fc = nn.Linear(int(1024 * alpha), class_num)
    # self.avg = nn.AdaptiveAvgPool2d(7)
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):
                kaiming_init(m)
            elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
                constant_init(m, 1)
def forward(self, x):
    outs=[]
    x = self.stem(x)
    x=self.conv1(x)
    outs.append(x)
    x = self.conv2(x)
    outs.append(x)
    x = self.conv3(x)
    outs.append(x)
    x = self.conv4(x)
    # x = self.avg(x)
    outs.append(x)
    # x = x.view(x.size(0), -1)
    #x = self.fc(x)
    return tuple(outs)
def train(self, mode=True):
    super(MobileNet, self).train(mode)
    if mode:
        for m in self.modules():
        # trick: eval have effect on BatchNorm only
            if isinstance(m, nn.BatchNorm2d):
                m.train()

3、init.py: from .mobilenet import MobileNet all = ['ResNet', 'make_res_layer', 'ResNeXt', 'SSDVGG', 'HRNet','MobileNet']

4、training command: python tools/train.py ./configs/foveabox/fovea_align_gn_ms_mobilenet_fpn_4gpu_2x.py --gpus 1 --work_dir ./checkpoints/mobilenet_Humancrowd

5、Error: File "/home/ubuntu/..../mmdet/models/backbones/mobilenet.py", line 181, in forward x = self.stem(x) File "/home/ubuntu/anaconda3/envs/torch-gpu/lib/python3.6/site-packages/torch/nn/modules/module.py", line 591, in getattr type(self).name, name)) AttributeError: 'MobileNet' object has no attribute 'stem'

I can run the training with resnet50. Is my mobilenet.py something wrong with mmdet?
could you show me your mobilenet.py and config ,please ?