MMDetection系列之（自定义模型）

我们基本上将模型组件分为5种类型。
backbone:通常是一个FCN网络提取特征地图，如ResNet, MobileNet。
neck:在脊骨和头部之间的部位，如:FPN, PAFPN…
head:用于特定任务的组件，如框预测、掩码预测等。
roi提取器:用于从特征映射中提取Rol特征的部分，如Rol Align。
loss:头部用于计算损耗的组件，如FocalLoss, L1Loss, GHMLoss。

backbone: usually an FCN network to extract feature maps, e.g., ResNet, MobileNet.

neck: the component between backbones and heads, e.g., FPN, PAFPN.

head: the component for specific tasks, e.g., bbox prediction and mask prediction.

roi extractor: the part for extracting RoI features from feature maps, e.g., RoI Align.

loss: the component in head for calculating losses, e.g., FocalLoss, L1Loss, and GHMLoss.

开发新的组件

Add a new backbone

这里我们将以MobileNet为例展示如何开发新组件。

定义一个新的主干(例如MobileNet)

Create a new file mmdet/models/backbones/mobilenet.py.

import torch.nn as nn

from ..builder import BACKBONES


@BACKBONES.register_module()
class MobileNet(nn.Module):

    def __init__(self, arg1, arg2):
        pass

    def forward(self, x):  # should return a tuple
        pass

2、Import the module

You can either add the following line to mmdet/models/backbones/init.py

from .mobilenet import MobileNet

or alternatively add

custom_imports = dict(
    imports=['mmdet.models.backbones.mobilenet'],
    allow_failed_imports=False)

配置文件，以避免修改原始代码。

3. Use the backbone in your config file

model = dict(
    ...
    backbone=dict(
        type='MobileNet',
        arg1=xxx,
        arg2=xxx),
    ...

Add new necks

1. Define a neck (e.g. PAFPN)

Create a new file mmdet/models/necks/pafpn.py.

from ..builder import NECKS

@NECKS.register_module()
class PAFPN(nn.Module):

    def __init__(self,
                in_channels,
                out_channels,
                num_outs,
                start_level=0,
                end_level=-1,
                add_extra_convs=False):
        pass

    def forward(self, inputs):
        # implementation is ignored
        pass

2. Import the module

You can either add the following line to mmdet/models/necks/init.py,

from .pafpn import PAFPN

or alternatively add

custom_imports = dict(
    imports=['mmdet.models.necks.pafpn.py'],
    allow_failed_imports=False)

to the config file and avoid modifying the original code.

3. Modify the config file

neck=dict(
    type='PAFPN',
    in_channels=[256, 512, 1024, 2048],
    out_channels=256,
    num_outs=5)

Add new heads

在这里，我们展示了如何开发一个新的头部与双头部R-CNN的例子如下。
首先，在mmdet/models/roi_heads/bbox_heads/double_bbox_head.py中添加一个新的box头。双头R-CNN实现了一种新的盒头用于目标检测。为了实现一个盒头，我们基本上需要实现新模块的三个功能，如下所示。

from mmdet.models.builder import HEADS
from .bbox_head import BBoxHead

@HEADS.register_module()
class DoubleConvFCBBoxHead(BBoxHead):
    r"""Bbox head used in Double-Head R-CNN

                                      /-> cls
                  /-> shared convs ->
                                      \-> reg
    roi features
                                      /-> cls
                  \-> shared fc    ->
                                      \-> reg
    """  # noqa: W605

    def __init__(self,
                 num_convs=0,
                 num_fcs=0,
                 conv_out_channels=1024,
                 fc_out_channels=1024,
                 conv_cfg=None,
                 norm_cfg=dict(type='BN'),
                 **kwargs):
        kwargs.setdefault('with_avg_pool', True)
        super(DoubleConvFCBBoxHead, self).__init__(**kwargs)


    def forward(self, x_cls, x_reg):

第二，如果有必要，实施一个新的Rol Head。我们计划从standard droihead继承新的DoubleHeadRoIHead。我们可以发现标准droihead已经实现了以下功能。

import torch

from mmdet.core import bbox2result, bbox2roi, build_assigner, build_sampler
from ..builder import HEADS, build_head, build_roi_extractor
from .base_roi_head import BaseRoIHead
from .test_mixins import BBoxTestMixin, MaskTestMixin


@HEADS.register_module()
class StandardRoIHead(BaseRoIHead, BBoxTestMixin, MaskTestMixin):
    """Simplest base roi head including one bbox head and one mask head.
    """

    def init_assigner_sampler(self):

    def init_bbox_head(self, bbox_roi_extractor, bbox_head):

    def init_mask_head(self, mask_roi_extractor, mask_head):


    def forward_dummy(self, x, proposals):


    def forward_train(self,
                      x,
                      img_metas,
                      proposal_list,
                      gt_bboxes,
                      gt_labels,
                      gt_bboxes_ignore=None,
                      gt_masks=None):

    def _bbox_forward(self, x, rois):

    def _bbox_forward_train(self, x, sampling_results, gt_bboxes, gt_labels,
                            img_metas):

    def _mask_forward_train(self, x, sampling_results, bbox_feats, gt_masks,
                            img_metas):

    def _mask_forward(self, x, rois=None, pos_inds=None, bbox_feats=None):


    def simple_test(self,
                    x,
                    proposal_list,
                    img_metas,
                    proposals=None,
                    rescale=False):
        """Test without augmentation."""

最后，用户需要在mmdet/models/bbox_heads/init.py和mmdet/models/roi_heads/init.py中添加模块，这样相应的注册表就可以找到并加载它们。

custom_imports=dict(
    imports=['mmdet.models.roi_heads.double_roi_head', 'mmdet.models.bbox_heads.double_bbox_head'])

到配置文件，并实现相同的目标。

_base_ = '../faster_rcnn/faster_rcnn_r50_fpn_1x_coco.py'
model = dict(
    roi_head=dict(
        type='DoubleHeadRoIHead',
        reg_roi_scale_factor=1.3,
        bbox_head=dict(
            _delete_=True,
            type='DoubleConvFCBBoxHead',
            num_convs=4,
            num_fcs=2,
            in_channels=256,
            conv_out_channels=1024,
            fc_out_channels=1024,
            roi_feat_size=7,
            num_classes=80,
            bbox_coder=dict(
                type='DeltaXYWHBBoxCoder',
                target_means=[0., 0., 0., 0.],
                target_stds=[0.1, 0.1, 0.2, 0.2]),
            reg_class_agnostic=False,
            loss_cls=dict(
                type='CrossEntropyLoss', use_sigmoid=False, loss_weight=2.0),
            loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=2.0))))

从MMDetection 2.0开始，配置系统支持继承配置，这样用户就可以专注于修改。DoubleHead R-CNN主要使用了一个新的DoubleHeadRolHead和一个新的doubeconvfcbboxhead，参数是根据每个模块的_init_函数来设置的。

Add new loss

假设您想添加一个新的损失作为MyLoss，用于边界框回归。要添加一个新的损失函数，用户需要在mmdet/models/losses/my_loss.py中实现它。The decorator weighted_loss enable the loss to be weighted for each element.

import torch
import torch.nn as nn

from ..builder import LOSSES
from .utils import weighted_loss

@weighted_loss
def my_loss(pred, target):
    assert pred.size() == target.size() and target.numel() > 0
    loss = torch.abs(pred - target)
    return loss

@LOSSES.register_module()
class MyLoss(nn.Module):

    def __init__(self, reduction='mean', loss_weight=1.0):
        super(MyLoss, self).__init__()
        self.reduction = reduction
        self.loss_weight = loss_weight

    def forward(self,
                pred,
                target,
                weight=None,
                avg_factor=None,
                reduction_override=None):
        assert reduction_override in (None, 'none', 'mean', 'sum')
        reduction = (
            reduction_override if reduction_override else self.reduction)
        loss_bbox = self.loss_weight * my_loss(
            pred, target, weight, reduction=reduction, avg_factor=avg_factor)
        return loss_bbox

Then the users need to add it in the mmdet/models/losses/init.py.

from .my_loss import MyLoss, my_loss

Alternatively, you can add

custom_imports=dict(
    imports=['mmdet.models.losses.my_loss'])

to the config file and achieve the same goal.
To use it, modify the loss_xxx field. Since MyLoss is for regression, you need to modify the loss_bbox field in the head.

oss_bbox=dict(type='MyLoss', loss_weight=1.0))

新模型使用问题