Customize Model
我们基本上将模型组件分为5种类型。
- backbone:
- neck:
- head:
- roi extractor:
- loss:
开发一个新组件
添加新的 backbone(以 MobileNet 为例)
定义新的 backbone
在
mmdet/models/backbones/mobilenet.py
创建一个新的文件mobilenet.py
import torch.nn as nn
from ..builder import BACKBONES
@BACKBONES.register_module()
class MobileNet(nn.Module):
def __init__(self, arg1, arg2):
# define the neccessary components (like pytorch)
pass
def forward(self, x): # should return a tuple
# using components to build your forward process
pass导入模块
在
mmdet/models/backbones/__init__.py
添加如下的语句:from .mobilenet import MobileNet
或者可以在配置文件里面添加,这样就不用修改原本的代码了。
custom_imports = dict(
imports=['mmdet.models.backbones.mobilenet'],
allow_failed_imports=False)在配置文件中使用新添加的 backbone
model = dict(
...
backbone=dict(
type='MobileNet',
arg1=xxx,
arg2=xxx),
...
添加新的 neck (以 PAFPN 为例)
定义新的 backbone
在
mmdet/models/backbones/mobilenet.py
创建一个新的文件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导入模块
在
mmdet/models/backbones/__init__.py
添加如下的语句:from .pafpn import PAFPN
或者可以在配置文件里面添加,这样就不用修改原本的代码了。
custom_imports = dict(
imports=['mmdet.models.necks.pafpn.py'],
allow_failed_imports=False)在配置文件中使用新添加的 neck
neck=dict(
type='PAFPN',
in_channels=[256, 512, 1024, 2048],
out_channels=256,
num_outs=5)
添加新的 heads
这里我们以双头R-CNN为例说明如何开发一个新的头,如下所示。
首先,在 mmdet/models/roi_heads/bbox_heads/double_box_head.py
中添加一个新的 bbox head。双头 R-CNN实现了一个新的 bbox head,用于物体检测。
为了实现一个 bbox head,基本上我们需要实现新模块的三个函数,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,实现一个新的 RoI Head
。我们从 StandardRoIHead 继承新的 DoubleHeadRoIHead。我们可以发现,一个 StandardRoIHead 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."""
Double Head 的修改主要体现在 bbox_forward 逻辑上,它继承了 StandardRoIHead 的其他逻辑。在 mmdet/models/roi_heads/double_roi_head.py
中,我们将新的 RoI Head 实现为如下。
from ..builder import HEADS
from .standard_roi_head import StandardRoIHead
@HEADS.register_module()
class DoubleHeadRoIHead(StandardRoIHead):
"""RoI head for Double Head RCNN
https://arxiv.org/abs/1904.06493
"""
def __init__(self, reg_roi_scale_factor, **kwargs):
super(DoubleHeadRoIHead, self).__init__(**kwargs)
self.reg_roi_scale_factor = reg_roi_scale_factor
def _bbox_forward(self, x, rois):
bbox_cls_feats = self.bbox_roi_extractor(
x[:self.bbox_roi_extractor.num_inputs], rois)
bbox_reg_feats = self.bbox_roi_extractor(
x[:self.bbox_roi_extractor.num_inputs],
rois,
roi_scale_factor=self.reg_roi_scale_factor)
if self.with_shared_head:
bbox_cls_feats = self.shared_head(bbox_cls_feats)
bbox_reg_feats = self.shared_head(bbox_reg_feats)
cls_score, bbox_pred = self.bbox_head(bbox_cls_feats, bbox_reg_feats)
bbox_results = dict(
cls_score=cls_score,
bbox_pred=bbox_pred,
bbox_feats=bbox_cls_feats)
return bbox_results
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,这样相应的注册表可以找到并加载它们,或者直接在配置文件添加如下语句:
custom_imports=dict(
imports=['mmdet.models.roi_heads.double_roi_head', 'mmdet.models.bbox_heads.double_bbox_head'])
然后在配置文件中使用新添加的组件
_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,这样用户可以专注于修改。双头 R-CNN 主要使用一个新的 DoubleHeadRoIHead 和一个新的 DoubleConvFCBBoxHead,参数根据每个模块的 __init__
函数来设置。
添加新的 loss
假设你想添加一个新的损失为 MyLoss,用于bounding box regression。为了增加一个新的损失函数,用户需要在 mmdet/models/losses/my_loss.py
中实现它。检测器的 weighted_loss 可以使损失对每个元素进行加权处理。
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
然后用户需要在 mmdet/models/losses/__init__.py
中添加代码:
from .my_loss import MyLoss, my_loss
或者直接在配置文件使用:
custom_imports=dict(
imports=['mmdet.models.losses.my_loss'])
要使用它,请修改 loss_xxx
字段。由于 MyLoss 是用于回归,你需要修改 head 中的 loss_bbox 字段。
loss_bbox=dict(type='MyLoss', loss_weight=1.0)