What version of wandb are you using?

On unlabeled data, we first
construct a candidate bag for each ground truth g by the traditional IoU-based
strategy, where the IoU threshold t is set to a relatively low value, e.g., 0.4
as default, to contain more proposals

does anyone has met this problem while training? 请问有和我一样在运行bash tools/dist_train_partially_labeled.sh的时候有这样报错的吗？

流程图看着就是 不同大小的图像进行训练，P3 和P2 特征怎么综合利用？

您好，想请教一下对于尺度不变性学习部分
1、原尺寸图用P3-P7，缩小一半的用P2-P6，请问推理预测eval的时候用哪一部分呢?
2、好像一般的ResNet50-FasterRCNN-FPN架构中P2—P5对应（C2-C5），P6P7是在P5上延伸出来的，请问PseCo中是这样吗，配置文件中的add_extra_convs='on_input'是这样的意思吗？

Can you tell me if there is a problem with this type of RP calculation? Because TP needs to ensure that a GTBox matches at most one prediction box, but as the above code TP is likely to be larger than the number of GTBoxes

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/A9dCpjHPfE or add me on WeChat (ID: 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.

What version of prettytable are you using, Im haveing problems installing prettytable

Collecting prettytable
Downloading prettytable-2.5.0-py3-none-any.whl (24 kB)
Requirement already satisfied: matplotlib>=2.1.0 in /ext3/miniconda3/envs/PseCo/lib/python3.6/site-packages (from pycocotools->-r requirements.txt (line 1)) (3.2.2)
Requirement already satisfied: numpy in /ext3/miniconda3/envs/PseCo/lib/python3.6/site-packages (from pycocotools->-r requirements.txt (line 1)) (1.18.5)
Requirement already satisfied: typing-extensions in /ext3/miniconda3/envs/PseCo/lib/python3.6/site-packages (from torch==1.9.0->-r requirements.txt (line 2)) (3.7.4.2)
Collecting dataclasses; python_version < "3.7"
Downloading dataclasses-0.8-py3-none-any.whl (19 kB)
Requirement already satisfied: pillow>=5.3.0 in /ext3/miniconda3/envs/PseCo/lib/python3.6/site-packages (from torchvision==0.10.0->-r requirements.txt (line 3)) (7.2.0)
Collecting addict
Downloading addict-2.4.0-py3-none-any.whl (3.8 kB)
Requirement already satisfied: pyyaml in /ext3/miniconda3/envs/PseCo/lib/python3.6/site-packages (from mmcv-full==1.3.9->-r requirements.txt (line 4)) (5.3.1)
Requirement already satisfied: yapf in /ext3/miniconda3/envs/PseCo/lib/python3.6/site-packages (from mmcv-full==1.3.9->-r requirements.txt (line 4)) (0.30.0)
Collecting opencv-python>=3
Downloading opencv_python-4.6.0.66-cp36-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (60.9 MB)
|################################| 60.9 MB 33.3 MB/s
Collecting GitPython==0.3.2.RC1
Downloading GitPython-0.3.2.RC1.tar.gz (313 kB)
|################################| 313 kB 100.7 MB/s
Collecting Parsley==1.2
Downloading Parsley-1.2.tar.gz (275 kB)
|################################| 275 kB 101.7 MB/s
ERROR: Command errored out with exit status 1:
command: /ext3/miniconda3/envs/PseCo/bin/python -c 'import sys, setuptools, tokenize; sys.argv[0] = '"'"'/tmp/pip-install-syt2zakr/Parsley/setup.py'"'"'; file='"'"'/tmp/pip-install-syt2zakr/Parsley/setup.py'"'"';f=getattr(tokenize, '"'"'open'"'"', open)(file);code=f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, file, '"'"'exec'"'"'))' egg_info --egg-base /tmp/pip-pip-egg-info-jjjsdcjj
cwd: /tmp/pip-install-syt2zakr/Parsley/
Complete output (7 lines):
Traceback (most recent call last):
File "", line 1, in
File "/tmp/pip-install-syt2zakr/Parsley/setup.py", line 16, in
long_description=open("README").read(),
File "/ext3/miniconda3/envs/PseCo/lib/python3.6/encodings/ascii.py", line 26, in decode
return codecs.ascii_decode(input, self.errors)[0]
UnicodeDecodeError: 'ascii' codec can't decode byte 0xe2 in position 3802: ordinal not in range(128)
----------------------------------------
ERROR: Command errored out with exit status 1: python setup.py egg_info Check the logs for full command output.
FATAL: While performing build: while running engine: exit status 1

使用mmdetection默认的方式，在config文件里加入load_from的coco预训练路径会报模型参数不匹配的错误，怎么才能正确将coco预训练模型参数加载到student和teacher model里呢

Hi, thanks for your such great work, the code is also very excellent ! But I have some problems when I read your code:
In your paper, you said you didn't do the bounding box regression on unlabeled data, but actually you did, I guess this is a writing mistake. What makes me confused is that you said you use the average of multiple positive proposals' IOU to estimate the reliability of a pseudo label, and reweight the regression loss. But, in your code, you calculate the reg_scale_factor to reweight the regression loss of the pseudo label, and the calculation of the reg_scale_factor is a little bit different from the paper, that is: the sample-wise reweighting without normalization in the paper, but reweight the whole regression loss with normalized reg_scale_factor in the code, so I want to ask the reason or the intuition behind this.

请问可以更换一些其他的检测模块吗？如cascade rcnn等

RuntimeError: Timed out initializing process group in store based barrier on rank: 0, for key: store_based_barrier_key:1 (world_size=4, worker_count=16, timeout=0:30:00)
ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: 1) local_rank: 0 (pid: 820989) of binary: /home/lihejun/anaconda3/envs/semi-det/bin/python
ERROR:torch.distributed.elastic.agent.server.local_elastic_agent:[default] Worker group failed
INFO:torch.distributed.elastic.agent.server.api:Local worker group finished (FAILED). Waiting 300 seconds for other agents to finish
/home/lihejun/anaconda3/envs/semi-det/lib/python3.6/site-packages/torch/distributed/elastic/utils/store.py:71: FutureWarning: This is an experimental API and will be changed in future.

label-level consistency learning，标签级别的一致性学习 和普通的训练什么区别呢？

您好，下面ppt中的proposal是如何产生的？是strong img aug-->student backbone-->student neck-->teacher head
还是strong img aug-->student backbone-->student neck-->student head ?

Hi, when I run the code (I only use the pretrained backbone resnet, and start training on my own dataset), I met this:

warnings.warn(
2022-09-13 15:15:48,002 - mmcv - INFO - load model from: open-mmlab://detectron2/resnet50_caffe
2022-09-13 15:15:48,002 - mmcv - INFO - Use load_from_openmmlab loader
2022-09-13 15:15:48,052 - mmcv - WARNING - The model and loaded state dict do not match exactly

unexpected key in source state_dict: conv1.bias

I don't know why and how to solve it? It seems like the progress can run, but if this will affect the performance or something else?

Looking forward to your reply!

Hello!

This is more of a question than an issue.

What kind of changes do I need to do to make this code work on the BDD100K dataset?

following #4
i'm using the mmdet provided the project. However, this error occurs.

Python: 3.6.13 |Anaconda, Inc.| (default, Jun  4 2021, 14:25:59) [GCC 7.5.0]                                                                                                                                     CUDA available: True                                                                                                                                                                                             GPU 0,1: Tesla PG503-216                                                                                                                                                                                         CUDA_HOME: /usr/local/cuda                                                                                                                                                                                       NVCC: Build cuda_11.2.r11.2/compiler.29618528_0
GCC: gcc (GCC) 5.4.0
PyTorch: 1.9.0
TorchVision: 0.10.0
OpenCV: 4.6.0
MMCV: 1.3.9
MMCV Compiler: GCC 7.3
MMCV CUDA Compiler: 10.2
MMDetection: 2.16.0+33b06ee

特征一致性学习 在训练的过程中怎么使用的呢？

Hello, I want to train my own data set at a ratio of 10%. At present, there are 7000 image data, of which 700 have been labeled and 6500 have not been labeled. How can I generate label.json and unlabel.json?

Can you upload model weights to google drive for easy download?

Trained model with default setting (default data separation, config: configs/PseCo/PseCo_faster_rcnn_r50_caffe_fpn_coco_180k.py) and I am getting mAP 19.6 for 1% labeled data and mAP 31.4 for 10% labeled data, against 22.43 for 1% and 36.06 for 10% reported in paper.

Also for each iteration, I keep getting "unsup_precision: 0.0000, unsup_recall: 0.0000". Does this look suspicious?

What could be the potential problems? Followed are some of my guesses:

1. Incompatible environment, which might mistaken some calculations
2. Non-optimal configs. In this case, may someone share an optimal configuration?

请问下作者为什么多分类任务要用二分类的FocalLoss

loading annotations into memory...
Traceback (most recent call last):
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/mmcv/utils/registry.py", line 52, in build_from_cfg
return obj_cls(**args)
File "/home/liuhaolin/PseCo/thirdparty/mmdetection/mmdet/datasets/custom.py", line 89, in init
self.data_infos = self.load_annotations(self.ann_file)
File "/home/liuhaolin/PseCo/thirdparty/mmdetection/mmdet/datasets/coco.py", line 49, in load_annotations
self.coco = COCO(ann_file)
File "/home/liuhaolin/PseCo/thirdparty/mmdetection/mmdet/datasets/api_wrappers/coco_api.py", line 28, in init
super().init(annotation_file=annotation_file)
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/pycocotools/coco.py", line 81, in init
with open(annotation_file, 'r') as f:
FileNotFoundError: [Errno 2] No such file or directory: '../data/annotations/semi_supervised/[email protected]'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/mmcv/utils/registry.py", line 52, in build_from_cfg
return obj_cls(**args)
File "/home/liuhaolin/PseCo/ssod/datasets/dataset_wrappers.py", line 10, in init
super().init([build_dataset(sup), build_dataset(unsup)], **kwargs)
File "/home/liuhaolin/PseCo/thirdparty/mmdetection/mmdet/datasets/builder.py", line 78, in build_dataset
dataset = build_from_cfg(cfg, DATASETS, default_args)
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/mmcv/utils/registry.py", line 55, in build_from_cfg
raise type(e)(f'{obj_cls.name}: {e}')
FileNotFoundError: CocoDataset: [Errno 2] No such file or directory: '../data/annotations/semi_supervised/[email protected]'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "tools/train.py", line 198, in
main()
File "tools/train.py", line 172, in main
datasets = [build_dataset(cfg.data.train)]
File "/home/liuhaolin/PseCo/thirdparty/mmdetection/mmdet/datasets/builder.py", line 78, in build_dataset
dataset = build_from_cfg(cfg, DATASETS, default_args)
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/mmcv/utils/registry.py", line 55, in build_from_cfg
raise type(e)(f'{obj_cls.name}: {e}')
FileNotFoundError: SemiDataset: CocoDataset: [Errno 2] No such file or directory: '../data/annotations/semi_supervised/[email protected]'
loading annotations into memory...
Traceback (most recent call last):
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/mmcv/utils/registry.py", line 52, in build_from_cfg
return obj_cls(**args)
File "/home/liuhaolin/PseCo/thirdparty/mmdetection/mmdet/datasets/custom.py", line 89, in init
self.data_infos = self.load_annotations(self.ann_file)
File "/home/liuhaolin/PseCo/thirdparty/mmdetection/mmdet/datasets/coco.py", line 49, in load_annotations
self.coco = COCO(ann_file)
File "/home/liuhaolin/PseCo/thirdparty/mmdetection/mmdet/datasets/api_wrappers/coco_api.py", line 28, in init
super().init(annotation_file=annotation_file)
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/pycocotools/coco.py", line 81, in init
with open(annotation_file, 'r') as f:
FileNotFoundError: [Errno 2] No such file or directory: '../data/annotations/semi_supervised/[email protected]'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/mmcv/utils/registry.py", line 52, in build_from_cfg
return obj_cls(**args)
File "/home/liuhaolin/PseCo/ssod/datasets/dataset_wrappers.py", line 10, in init
super().init([build_dataset(sup), build_dataset(unsup)], **kwargs)
File "/home/liuhaolin/PseCo/thirdparty/mmdetection/mmdet/datasets/builder.py", line 78, in build_dataset
dataset = build_from_cfg(cfg, DATASETS, default_args)
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/mmcv/utils/registry.py", line 55, in build_from_cfg
raise type(e)(f'{obj_cls.name}: {e}')
FileNotFoundError: CocoDataset: [Errno 2] No such file or directory: '../data/annotations/semi_supervised/[email protected]'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "tools/train.py", line 198, in
main()
File "tools/train.py", line 172, in main
datasets = [build_dataset(cfg.data.train)]
File "/home/liuhaolin/PseCo/thirdparty/mmdetection/mmdet/datasets/builder.py", line 78, in build_dataset
dataset = build_from_cfg(cfg, DATASETS, default_args)
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/mmcv/utils/registry.py", line 55, in build_from_cfg
raise type(e)(f'{obj_cls.name}: {e}')
FileNotFoundError: SemiDataset: CocoDataset: [Errno 2] No such file or directory: '../data/annotations/semi_supervised/[email protected]'
Killing subprocess 94576
Killing subprocess 94577
Traceback (most recent call last):
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/runpy.py", line 194, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/runpy.py", line 87, in _run_code
exec(code, run_globals)
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/torch/distributed/launch.py", line 340, in
main()
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/torch/distributed/launch.py", line 326, in main
sigkill_handler(signal.SIGTERM, None) # not coming back
File "/sdb/liuhaolin/anaconda3/envs/pseco/lib/python3.8/site-packages/torch/distributed/launch.py", line 301, in sigkill_handler
raise subprocess.CalledProcessError(returncode=last_return_code, cmd=cmd)
subprocess.CalledProcessError: Command '['/sdb/liuhaolin/anaconda3/envs/pseco/bin/python', '-u', 'tools/train.py', '--local_rank=1', 'configs/PseCo/PseCo_faster_rcnn_r50_caffe_fpn_coco_180k.py', '--work-dir', '/home/liuhaolin/PseCo/', '--launcher=pytorch', '--cfg-options', 'fold=1', 'percent=10']' returned non-zero exit status 1.

how to install this version? i only found 2.16.0.
what's the meaning of fe46ffe

Hi, when will you release the code for the one stage detector?
And would you plan to support RetinaNet?

Are there plans to adapt the framework to other network structures, like htc++ in softteacher?

2022-11-13 14:52:08,974 - mmdet.ssod - INFO - Clone all parameters of student to teacher...
2022-11-13 14:53:37,110 - mmdet.ssod - INFO - Iter [50/180000] lr: 9.890e-04, eta: 3 days, 15:47:15, time: 1.756, data_time: 0.067, memory: 5337, ema_momentum: 0.9800, unsup_loss_rpn_cls: 0.0000, unsup_loss_rpn_bbox: 0.0000, unsup_loss_rpn_cls_V2: 0.0000, unsup_loss_rpn_bbox_V2: 0.0000, unsup_loss_cls: 0.0000, unsup_loss_bbox: 0.0000, unsup_loss_cls_V2: 0.0000, unsup_loss_bbox_V2: 0.0000, unsup_tea_pos_score_mean: 0.0000, unsup_tea_pos_score_min: 0.0000, unsup_cls_score_thr: 0.0000, unsup_pos_number: 0.0000, unsup_precision: 0.0000, unsup_recall: 0.0000, loss: 0.0000
2022-11-13 14:56:06,218 - mmdet.ssod - INFO - Iter [100/180000] lr: 1.988e-03, eta: 4 days, 22:23:38, time: 2.982, data_time: 0.045, memory: 5449, ema_momentum: 0.9900, unsup_loss_rpn_cls: 0.0000, unsup_loss_rpn_bbox: 0.0000, unsup_loss_rpn_cls_V2: 0.0000, unsup_loss_rpn_bbox_V2: 0.0000, unsup_loss_cls: 0.0000, unsup_loss_bbox: 0.0000, unsup_loss_cls_V2: 0.0000, unsup_loss_bbox_V2: 0.0000, unsup_tea_pos_score_mean: 0.0000, unsup_tea_pos_score_min: 0.0000, unsup_cls_score_thr: 0.0000, unsup_pos_number: 0.0000, unsup_precision: 0.0000, unsup_recall: 0.0000, loss: 0.0000
2022-11-13 14:58:55,606 - mmdet.ssod - INFO - Iter [150/180000] lr: 2.987e-03, eta: 5 days, 15:19:17, time: 3.388, data_time: 0.043, memory: 5449, ema_momentum: 0.9933, unsup_loss_rpn_cls: 0.0000, unsup_loss_rpn_bbox: 0.0000, unsup_loss_rpn_cls_V2: 0.0000, unsup_loss_rpn_bbox_V2: 0.0000, unsup_loss_cls: 0.0000, unsup_loss_bbox: 0.0000, unsup_loss_cls_V2: 0.0000, unsup_loss_bbox_V2: 0.0000, unsup_tea_pos_score_mean: 0.0000, unsup_tea_pos_score_min: 0.0000, unsup_cls_score_thr: 0.0000, unsup_pos_number: 0.0000, unsup_precision: 0.0000, unsup_recall: 0.0000, loss: 0.0000
2022-11-13 15:01:45,491 - mmdet.ssod - INFO - Iter [200/180000] lr: 3.986e-03, eta: 5 days, 23:53:15, time: 3.398, data_time: 0.044, memory: 5552, ema_momentum: 0.9950, unsup_loss_rpn_cls: 0.0000, unsup_loss_rpn_bbox: 0.0000, unsup_loss_rpn_cls_V2: 0.0000, unsup_loss_rpn_bbox_V2: 0.0000, unsup_loss_cls: 0.0000, unsup_loss_bbox: 0.0000, unsup_loss_cls_V2: 0.0000, unsup_loss_bbox_V2: 0.0000, unsup_tea_pos_score_mean: 0.0000, unsup_tea_pos_score_min: 0.0000, unsup_cls_score_thr: 0.0000, unsup_pos_number: 0.0000, unsup_precision: 0.0000, unsup_recall: 0.0000, loss: 0.0000
2022-11-13 15:03:23,273 - mmdet.ssod - INFO - Iter [250/180000] lr: 4.985e-03, eta: 5 days, 14:36:26, time: 1.956, data_time: 0.043, memory: 5552, ema_momentum: 0.9960, unsup_loss_rpn_cls: 0.0000, unsup_loss_rpn_bbox: 0.0000, unsup_loss_rpn_cls_V2: 0.0000, unsup_loss_rpn_bbox_V2: 0.0000, unsup_loss_cls: 0.0000, unsup_loss_bbox: 0.0000, unsup_loss_cls_V2: 0.0000, unsup_loss_bbox_V2: 0.0000, unsup_tea_pos_score_mean: 0.0000, unsup_tea_pos_score_min: 0.0000, unsup_cls_score_thr: 0.0000, unsup_pos_number: 0.0000, unsup_precision: 0.0000, unsup_recall: 0.0000, loss: 0.0000
2022-11-13 15:04:58,171 - mmdet.ssod - INFO - Iter [300/180000] lr: 5.984e-03, eta: 5 days, 7:55:52, time: 1.898, data_time: 0.043, memory: 5552, ema_momentum: 0.9967, unsup_loss_rpn_cls: 0.0000, unsup_loss_rpn_bbox: 0.0000, unsup_loss_rpn_cls_V2: 0.0000, unsup_loss_rpn_bbox_V2: 0.0000, unsup_loss_cls: 0.0000, unsup_loss_bbox: 0.0000, unsup_loss_cls_V2: 0.0000, unsup_loss_bbox_V2: 0.0000, unsup_tea_pos_score_mean: 0.0000, unsup_tea_pos_score_min: 0.0000, unsup_cls_score_thr: 0.0000, unsup_pos_number: 0.0000, unsup_precision: 0.0000, unsup_recall: 0.0000, loss: 0.0000
2022-11-13 15:07:47,249 - mmdet.ssod - INFO - Iter [350/180000] lr: 6.983e-03, eta: 5 days, 13:43:52, time: 3.381, data_time: 0.042, memory: 5552, ema_momentum: 0.9971, unsup_loss_rpn_cls: 0.0000, unsup_loss_rpn_bbox: 0.0000, unsup_loss_rpn_cls_V2: 0.0000, unsup_loss_rpn_bbox_V2: 0.0000, unsup_loss_cls: 0.0000, unsup_loss_bbox: 0.0000, unsup_loss_cls_V2: 0.0000, unsup_loss_bbox_V2: 0.0000, unsup_tea_pos_score_mean: 0.0000, unsup_tea_pos_score_min: 0.0000, unsup_cls_score_thr: 0.0000, unsup_pos_number: 0.0000, unsup_precision: 0.0000, unsup_recall: 0.0000, loss: 0.0000
2022-11-13 15:10:36,283 - mmdet.ssod - INFO - Iter [400/180000] lr: 7.982e-03, eta: 5 days, 18:03:53, time: 3.381, data_time: 0.045, memory: 5553, ema_momentum: 0.9975, unsup_loss_rpn_cls: 0.0000, unsup_loss_rpn_bbox: 0.0000, unsup_loss_rpn_cls_V2: 0.0000, unsup_loss_rpn_bbox_V2: 0.0000, unsup_loss_cls: 0.0000, unsup_loss_bbox: 0.0000, unsup_loss_cls_V2: 0.0000, unsup_loss_bbox_V2: 0.0000, unsup_tea_pos_score_mean: 0.0000, unsup_tea_pos_score_min: 0.0000, unsup_cls_score_thr: 0.0000, unsup_pos_number: 0.0000, unsup_precision: 0.0000, unsup_recall: 0.0000, loss: 0.0000