or I will try to merge your code into mmdetection

您好，看完您的paper覺得挺有趣的 因此想作一點實驗
但發現您的config檔按裡沒有loss ，想請問您當初loss怎麼弄得?

Great works! And I have several questions about your work.
The Adaptive Channel Fusion module seems to be the attention mechanism, right? And I'm wondering how to visualize the attention map in order to show its effectiveness.

where is the residual feature augmentation code? I can't find it in .py. @Gus-Guo

Could you please point out where the Residual Feature Augmentation module is?

Excuse me，Which version of MMDetction do you use？

I have some problems during the installation. Where is the 0.2.16 version of mmcv. It was not found in the website you gave.

Welcome update to OpenMMLab 2.0

I am Vansin, the technical operator of OpenMMLab. In September of last year, we announced the release of OpenMMLab 2.0 at the World Artificial Intelligence Conference in Shanghai. We invite you to upgrade your algorithm library to OpenMMLab 2.0 using MMEngine, which can be used for both research and commercial purposes. If you have any questions, please feel free to join us on the OpenMMLab Discord at https://discord.gg/amFNsyUBvm or add me on WeChat (van-sin) and I will invite you to the OpenMMLab WeChat group.

Here are the OpenMMLab 2.0 repos branches:

Attention: please create a new virtual environment for OpenMMLab 2.0.

在安装运行 cd AugFPN
bash compile.sh该步骤的时候
出现错误。
希望得到解答！！谢谢！！
^
In file included from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/ATen/core/Tensor.h:3:0,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/ATen/DeviceGuard.h:4,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/ATen/ATen.h:11,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/types.h:3,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader_options.h:4,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/base.h:3,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/stateful.h:3,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader.h:3,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/data.h:3,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/all.h:8,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/extension.h:4,
from src/deform_conv_cuda.cpp:4:
/root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/ATen/core/TensorBody.h:210:30: note: declared here
DeprecatedTypeProperties & type() const {
^~~~
src/deform_conv_cuda.cpp:605:34: warning: ‘at::DeprecatedTypeProperties& at::Tensor::type() const’ is deprecated: Tensor.type() is deprecated. Instead use Tensor.options(), which in many cases (e.g. in a constructor) is a drop-in replacement. If you were using data from type(), that is now available from Tensor itself, so instead of tensor.type().scalar_type(), use tensor.scalar_type() instead and instead of tensor.type().backend() use tensor.device(). [-Wdeprecated-declarations]
input.type());
^
In file included from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/ATen/core/Tensor.h:3:0,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/ATen/DeviceGuard.h:4,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/ATen/ATen.h:11,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/types.h:3,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader_options.h:4,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/base.h:3,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader/stateful.h:3,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/data/dataloader.h:3,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/data.h:3,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/csrc/api/include/torch/all.h:8,
from /root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/torch/extension.h:4,
from src/deform_conv_cuda.cpp:4:
/root/miniconda3/envs/open-mmlab-l/lib/python3.7/site-packages/torch/include/ATen/core/TensorBody.h:210:30: note: declared here
DeprecatedTypeProperties & type() const {
^~~~
error: command '/usr/bin/gcc' failed with exit code 1

I read this part of the paper many times, but I couldn't understand what you want to say. what is the ratio? why four figures both have 0,1,2,3 in the horizontal axis? Looking forward to your reply!

@Gus-Guo

Traceback (most recent call last):
File "setup.py", line 15, in
cmdclass={'build_ext': BuildExtension})
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/site-packages/setuptools/init.py", line 161, in setup
return distutils.core.setup(**attrs)
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/distutils/core.py", line 148, in setup
dist.run_commands()
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/distutils/dist.py", line 966, in run_commands
self.run_command(cmd)
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/distutils/dist.py", line 985, in run_command
cmd_obj.run()
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/site-packages/setuptools/command/build_ext.py", line 87, in run
_build_ext.run(self)
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/site-packages/Cython/Distutils/old_build_ext.py", line 186, in run
_build_ext.build_ext.run(self)
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/distutils/command/build_ext.py", line 340, in run
self.build_extensions()
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/site-packages/torch/utils/cpp_extension.py", line 580, in build_extensions
build_ext.build_extensions(self)
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/site-packages/Cython/Distutils/old_build_ext.py", line 195, in build_extensions
_build_ext.build_ext.build_extensions(self)
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/distutils/command/build_ext.py", line 449, in build_extensions
self._build_extensions_serial()
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/distutils/command/build_ext.py", line 474, in _build_extensions_serial
self.build_extension(ext)
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/site-packages/setuptools/command/build_ext.py", line 208, in build_extension
_build_ext.build_extension(self, ext)
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/distutils/command/build_ext.py", line 534, in build_extension
depends=ext.depends)
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/site-packages/torch/utils/cpp_extension.py", line 423, in unix_wrap_ninja_compile
with_cuda=with_cuda)
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/site-packages/torch/utils/cpp_extension.py", line 1140, in _write_ninja_file_and_compile_objects
error_prefix='Error compiling objects for extension')
File "/root/anaconda3/envs/clj_mmdet/lib/python3.7/site-packages/torch/utils/cpp_extension.py", line 1413, in _run_ninja_build
raise RuntimeError(message)
RuntimeError: Error compiling objects for extension

您好，打扰了，训练时突然出现这个问题，请教一下，应该怎么解决

When i test the trained module.i input this conmmad

python tools/test.py configs/faster_rcnn_r50_augfpn_1x.py epoch_8.pth --gpus 4 --out results.pkl --eval bbox

but it occured an error:
NoneType' object has no attribute 'rcnn'

deteil:

File "tools/test.py", line 127, in <module>
    main()
  File "tools/test.py", line 78, in main
    cfg.model, train_cfg=None, test_cfg=cfg.test_cfg)
  File "/home/superfly/AugFPN-master/mmdet/models/builder.py", line 51, in build_detector
    return build(cfg, DETECTORS, dict(train_cfg=train_cfg, test_cfg=test_cfg))
  File "/home/superfly/AugFPN-master/mmdet/models/builder.py", line 31, in build
    return _build_module(cfg, registry, default_args)
  File "/home/superfly/AugFPN-master/mmdet/models/builder.py", line 23, in _build_module
    return obj_type(**args)
  File "/home/superfly/AugFPN-master/mmdet/models/detectors/faster_rcnn.py", line 25, in __init__
    pretrained=pretrained)
  File "/home/superfly/AugFPN-master/mmdet/models/detectors/two_stage.py", line 44, in __init__
    if 'use_consistent_supervision' in self.train_cfg.rcnn:
AttributeError: 'NoneType' object has no attribute 'rcnn'

Could you please point out where the Consistent Supervision and Soft RoI Selection are?

(hit2) htu@htu-System-Product-Name:~/bailu/AugFPN-master/mmcv-master$ /home/htu/bailu/AugFPN-master/tools/dist_train.sh configs/faster_rcnn_r50_augfpn_1x.py 1 --validate --work_dir faster_rcnn_r50_augfpn_1x
Traceback (most recent call last):
File "/home/htu/bailu/AugFPN-master/tools/train.py", line 7, in
from mmdet.datasets import get_datasets
ImportError: cannot import name 'get_datasets'
Traceback (most recent call last):
File "/home/htu/dongxu/anaconda3/envs/hit2/lib/python3.6/runpy.py", line 193, in _run_module_as_main
"main", mod_spec)
File "/home/htu/dongxu/anaconda3/envs/hit2/lib/python3.6/runpy.py", line 85, in _run_code
exec(code, run_globals)
File "/home/htu/dongxu/anaconda3/envs/hit2/lib/python3.6/site-packages/torch/distributed/launch.py", line 263, in
main()
File "/home/htu/dongxu/anaconda3/envs/hit2/lib/python3.6/site-packages/torch/distributed/launch.py", line 259, in main
cmd=cmd)
subprocess.CalledProcessError: Command '['/home/htu/dongxu/anaconda3/envs/hit2/bin/python', '-u', '/home/htu/bailu/AugFPN-master/tools/train.py', '--local_rank=0', 'configs/faster_rcnn_r50_augfpn_1x.py', '--launcher', 'pytorch', '--validate', '--work_dir', 'faster_rcnn_r50_augfpn_1x']' returned non-zero exit status 1.

您好，我在运行时出现这个问题，请问是环境配置出问题了吗：
loading annotations into memory...
Done (t=0.21s)
creating index...
index created!
Traceback (most recent call last):
File "tools/train.py", line 90, in
main()
File "tools/train.py", line 86, in main
logger=logger)
File "/home/linsy/anaconda3/envs/augfpn/lib/python3.7/site-packages/mmdet/apis/train.py", line 59, in train_detector
_non_dist_train(model, dataset, cfg, validate=validate)
File "/home/linsy/anaconda3/envs/augfpn/lib/python3.7/site-packages/mmdet/apis/train.py", line 113, in _non_dist_train
cfg.log_level)
File "/home/linsy/anaconda3/envs/augfpn/lib/python3.7/site-packages/mmcv-0.2.16-py3.7-linux-x86_64.egg/mmcv/runner/runner.py", line 46, in init
self.optimizer = self.init_optimizer(optimizer)
File "/home/linsy/anaconda3/envs/augfpn/lib/python3.7/site-packages/mmcv-0.2.16-py3.7-linux-x86_64.egg/mmcv/runner/runner.py", line 145, in init_optimizer
dict(params=self.model.parameters()))
File "/home/linsy/anaconda3/envs/augfpn/lib/python3.7/site-packages/mmcv-0.2.16-py3.7-linux-x86_64.egg/mmcv/runner/utils.py", line 50, in obj_from_dict
return obj_type(**args)
TypeError: init() got an unexpected keyword argument 'paramwise_options'

How can i apply it mmd2.0？

Hi~
May I ask why you set the α1,α2 and α3 in the “we first produce multiple context features with different scales of (α1 ×S, α2 ×S, .., αn ×S) by performing ratio-invariant adaptive pooling on C5 whose scale is S” as 0.1, 0.2, and 0.3? why not 0.3, 0.5, 0.8?
Looking forward to your reply!

AdapPool is not defined in class HighFPNRetinanet