Implementing YOLOv3 architecture #1695

wuhy08 · 2019-11-21T00:57:49Z

Hi MMlab,

I am Haoyu Wu, an employee at Western Digital. I have implemented the YOLOv3 architecture in the mmdetection framework. Hope that is helpful to community. It is noted that the contribution comes from my work at Western Digital, therefore each files have a header of copyright claims. But we agree to use Apache 2.0 license.

I have pretrained weight file. I wonder what is the best way to share that.

Thank you!

zxz-cc · 2020-01-15T03:28:58Z

Hi MMlab,

I am Haoyu Wu, an employee at Western Digital. I have implemented the YOLOv3 architecture in the mmdetection framework. Hope that is helpful to community. It is noted that the contribution comes from my work at Western Digital, therefore each files have a header of copyright claims. But we agree to use Apache 2.0 license.

I have pretrained weight file. I wonder what is the best way to share that.

Thank you!

Where is the pretrained weight file??

liugl7 · 2020-01-18T04:38:21Z

please give the download link to "darknet_state_dict_only.pth"

liugl7 · 2020-01-18T04:41:35Z

Hi MMlab,

I am Haoyu Wu, an employee at Western Digital. I have implemented the YOLOv3 architecture in the mmdetection framework. Hope that is helpful to community. It is noted that the contribution comes from my work at Western Digital, therefore each files have a header of copyright claims. But we agree to use Apache 2.0 license.

I have pretrained weight file. I wonder what is the best way to share that.

Thank you!

please give the download link to "darknet_state_dict_only.pth"
thank you very much !

ElectronicElephant · 2020-02-14T07:53:45Z

@wuhy08 Hello, great thanks to your contribution. Would you please offer a download link to the pretrained model?

ElectronicElephant · 2020-02-19T11:26:20Z

@CRmost @liugl7
You can try

https://drive.google.com/file/d/1-59qCzvSOCvVFCQ9HDggHA8xUQFR4LDx/view?usp=sharing

, which is extracted from the author's another repo.

If you get things right, please let me known. Thank you.

xvjiarui · 2020-03-11T12:40:44Z

Hi @wuhy08
Thanks for your contribution and sorry for the delay.
Could you please clean up the comment in the config files?

Clean the comments in config file

wuhy08 · 2020-03-12T02:29:36Z

Hi @xvjiarui

I cleaned the comments in the config file. The init files have some conflicts. I wonder how to deal with that.

xvjiarui · 2020-03-12T02:55:40Z

Hi @wuhy08
You could try git merge open-mmlab master to solve conflicts from your side.

xvjiarui · 2020-03-12T03:01:39Z

@wuhy08
Also, it would be nice if I could have the link to your pre-trained weight and trained model so that I could work on reproducing.
READMD.md is suggested to included under the configs/yolo, where you could also provide your links and results.

xvjiarui · 2020-03-12T07:07:23Z

Hi @wuhy08
FYI, ConvModule is just refactored in #2239, which could be used to replace ConvLayer in your implementation.

wuhy08 · 2020-03-12T16:35:19Z

@xvjiarui

I will try to merge your current master to my branch, resolve any conflict, and test it. It might take a few days. I will let you know by then.

# Conflicts: # mmdet/models/anchor_heads/__init__.py # mmdet/models/necks/__init__.py # mmdet/models/utils/__init__.py

GreenTeaHua · 2020-03-22T08:01:15Z

everything is ok？

wuhy08 · 2020-03-26T21:13:50Z

Hi @xvjiarui

Sorry for the delay due to coronavirus. But I pushed more commits to get it streamlined. I haven't use your new ConvModule since it require substantial amount of work including retraining/state_dict converting. So I will leave it for other contributors. I also uploaded my training checkpoint to Google Drive, linked in README. I hope that can be stored in your centralized storage in case it is unintentionally deleted. I am not authorized to merge so I think some conflicts have to be resolved by authorized merger.

Thank you and stay healthy!

miaoshuyu · 2020-04-30T02:42:56Z

@wuhy08
hi, I run your code and miss this problem, do you know why.
RuntimeError: Expected to have finished reduction in the prior iteration before starting a new one. This error indicates that your module has parameters that were not used in producing its output (the return value of forward). You can enable unused parameter detection by passing the keyword argument find_unused_parameters=True to torch.nn.parallel.DistributedDataParallel. If you already have this argument set, then the distributed data parallel module wasn't able to locate the output tensors in the return value of your module's forward function.

CLAassistant · 2020-05-21T15:55:15Z

All committers have signed the CLA.

hasanirtiza · 2020-06-06T09:54:12Z

@miaoshuyu
If you have not solved it already, at line 132 of mmdet/apis/train.py, when he is calling
model = MMDistributedDataParallel(model.cuda(), device_ids=[torch.cuda.current_device()], broadcast_buffers=False])

add an additional argument find_unused_parameters=True.

model = MMDistributedDataParallel(model.cuda(), device_ids=[torch.cuda.current_device()], broadcast_buffers=False, find_unused_parameters=True)

@LMerCy

* Implement YOLOv3 * Remove unused function * Update yolov3_ms_aug_273e.py Clean the comments in config file * Add README.md * port to mmdet-2.0 api * unify registry * port to ConvModule and remove ConvLayer * Refactor Backbone * Update README * Lint and format * Unify the class name * fix the `label - 1` problem * Move a lot hard-coded params to the __init__ function * Refactor YOLOV3Neck * Add norm_cfg and act_cfg to backbone * Update Config * Fix doc string * Fix nms (thanks to @LMerCy) * Add doc string * Update config * Remove pretrained in head and neck * Add support for conv_cfg in neck * Update mmdet/models/dense_heads/yolo_head.py Co-authored-by: Jerry Jiarui XU <xvjiarui0826@gmail.com> * Update mmdet/models/dense_heads/yolo_head.py Co-authored-by: Jerry Jiarui XU <xvjiarui0826@gmail.com> * Fix README.md * Fix typos * update config * flake8, yapf, docformatter, etc * Update README * Add conv_cfg to backbone and head * Move some config to arch_settings in backbone * Add doc strings and replace Warning with warnings.warn() * Fix bug. * Update doc * Add _frozen_stages for backbone * Update mmdet/models/backbones/darknet.py Co-authored-by: Jerry Jiarui XU <xvjiarui0826@gmail.com> * Fix inplace bug * fix indent * refactor config * set 8GPU lr * fixed typo * update performance table * Resolve conversation * Add anchor generator and coder * fixed test * Finish refactor * refactor anchor order * fixed batch size * Fixed train_cfg * fix yolo assigner * clean up * Fixed format * Update model zoo * change to mmcv pretrain link * add test forward * fixed comma and docstring * Refactor loss * reformat * fixed avg_factor * revert to original * fixed format * update table * fixed BCE Co-authored-by: Haoyu Wu <haoyu.wu@wdc.com> Co-authored-by: Haoyu Wu <wuhy08@users.noreply.github.com> Co-authored-by: Haoyu Wu <wuhaoyu1989@gmail.com> Co-authored-by: Jerry Jiarui XU <xvjiarui0826@gmail.com> Co-authored-by: xmpeng <1051323399@qq.com>

ZwwWayne · 2020-10-14T01:26:54Z

Since YOLOv3 is supported by #3083. This PR is closed. Thanks for your contribution.

OpenMMLab-Assistant-004 · 2023-04-12T04:25:49Z

Hi @wuhy08 ！We are grateful for your efforts in helping improve mmdetection open-source project during your personal time.

Welcome to join OpenMMLab Special Interest Group (SIG) private channel on Discord, where you can share your experiences, ideas, and build connections with like-minded peers. To join the SIG channel, simply message moderator— OpenMMLab on Discord or briefly share your open-source contributions in the #introductions channel and we will assist you. Look forward to seeing you there! Join us ：https://discord.gg/UjgXkPWNqA
If you have a WeChat account，welcome to join our community on WeChat. You can add our assistant ：openmmlabwx. Please add "mmsig + Github ID" as a remark when adding friends：）

Thank you again for your contribution❤

Haoyu Wu and others added 3 commits November 18, 2019 16:32

Implement YOLOv3

9b5b39e

Remove unused function

787def0

Merge branch 'master' into yolo

589158c

hellock mentioned this pull request Jan 8, 2020

Will yolov3 be added in the future? #1945

Closed

ElectronicElephant mentioned this pull request Feb 18, 2020

Question about the progress on #1695 #2110

Closed

xvjiarui self-requested a review March 9, 2020 11:32

Update yolov3_ms_aug_273e.py

722fef8

Clean the comments in config file

Merge remote-tracking branch 'origin-mm/master' into yolo

08ef678

# Conflicts: # mmdet/models/anchor_heads/__init__.py # mmdet/models/necks/__init__.py # mmdet/models/utils/__init__.py

Add README.md

d61c575

This was referenced Jun 18, 2020

Roadmap of MMDetection #2931

Open

YOLOv3 - Continue on #1695 #3083

Merged

ZwwWayne closed this Oct 14, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Implementing YOLOv3 architecture #1695

Implementing YOLOv3 architecture #1695

wuhy08 commented Nov 21, 2019

zxz-cc commented Jan 15, 2020

liugl7 commented Jan 18, 2020

liugl7 commented Jan 18, 2020

ElectronicElephant commented Feb 14, 2020

ElectronicElephant commented Feb 19, 2020 •

edited

Loading

xvjiarui commented Mar 11, 2020

wuhy08 commented Mar 12, 2020

xvjiarui commented Mar 12, 2020

xvjiarui commented Mar 12, 2020

xvjiarui commented Mar 12, 2020

wuhy08 commented Mar 12, 2020 •

edited

Loading

GreenTeaHua commented Mar 22, 2020

wuhy08 commented Mar 26, 2020

miaoshuyu commented Apr 30, 2020

CLAassistant commented May 21, 2020 •

edited

Loading

hasanirtiza commented Jun 6, 2020

ZwwWayne commented Oct 14, 2020

OpenMMLab-Assistant-004 commented Apr 12, 2023

Implementing YOLOv3 architecture #1695

Implementing YOLOv3 architecture #1695

Conversation

wuhy08 commented Nov 21, 2019

zxz-cc commented Jan 15, 2020

liugl7 commented Jan 18, 2020

liugl7 commented Jan 18, 2020

ElectronicElephant commented Feb 14, 2020

ElectronicElephant commented Feb 19, 2020 • edited Loading

xvjiarui commented Mar 11, 2020

wuhy08 commented Mar 12, 2020

xvjiarui commented Mar 12, 2020

xvjiarui commented Mar 12, 2020

xvjiarui commented Mar 12, 2020

wuhy08 commented Mar 12, 2020 • edited Loading

GreenTeaHua commented Mar 22, 2020

wuhy08 commented Mar 26, 2020

miaoshuyu commented Apr 30, 2020

CLAassistant commented May 21, 2020 • edited Loading

hasanirtiza commented Jun 6, 2020

ZwwWayne commented Oct 14, 2020

OpenMMLab-Assistant-004 commented Apr 12, 2023

ElectronicElephant commented Feb 19, 2020 •

edited

Loading

wuhy08 commented Mar 12, 2020 •

edited

Loading

CLAassistant commented May 21, 2020 •

edited

Loading