[Feature] Support DDRNet #1722

qyyyyq · 2022-06-29T09:08:20Z

Hi. I'm a beginner and I want to try to submit an implementation of DDRNet

Motivation

Support DDRNet
Paper: Deep Dual-resolution Networks for Real-time and Accurate Semantic Segmentation of Road Scenes
official Code: https://github.com/ydhongHIT/DDRNet

Modification

Added backbone code.
Added config file for cityscapes.
Added tools for converting pretrained models.

Note

1 The training tricks in the paper are not used.Only the basic configuration of mmsegmentation is used.
2 Pretrained models(ImageNet),including the official pretrained weight and the converted pretrained weight can be downloaded from https://drive.google.com/drive/folders/1By_LJCoZGN98i-JsP8xv22VnhIvZnm3w

CLAassistant · 2022-06-29T09:08:28Z

All committers have signed the CLA.

MengzhangLI · 2022-06-29T09:24:52Z

Hi, thanks for your nice PR. We have planned to support DDRNet and it is very helpful for this repo.

First, could you please run normally on those models with provided ckpts?

From DDRNet repo ddrnet-23-slim_in1k-pre_2x8_1024x1024_160k_cityscapes.py should be 77.4, ddrnet-23_in1k-pre_2x8_1024x1024_160k_cityscapes.py should be 79.4 and ddrnet-39_in1k-pre_2x8_1024x1024_160k_cityscapes.py should be 80.4.

Thanks in advance!

MengzhangLI · 2022-06-29T09:27:18Z

configs/ddrnet/ddrnet-23-slim_in1k-pre_2x8_1024x1024_160k_cityscapes.py

+    samples_per_gpu=8,
+    workers_per_gpu=8,


Suggested change

samples_per_gpu=8,

workers_per_gpu=8,

samples_per_gpu=4,

workers_per_gpu=4,

To keep the same with former Cityscapes dataset config, we may set default setting 4 GPUs with 4 samplers per GPU.

Sorry, i just evaluate models on the Cityscapes val set. DDRNet-23-slim gains 77.37% mIoU ,and DDRNet-23 gains 78.66% mIoU .Compared with val results (77.8%,79.5%) from DDRNet repo ,i think results are acceptable,because some basic training tricks such as OHEM are not used.DDRNet-39 gains 79.46% mIoU on Cityscapes val set.

Sorry, i just evaluate models on the Cityscapes val set. DDRNet-23-slim gains 77.37% mIoU ,and DDRNet-23 gains 78.66% mIoU .Compared with val results (77.8%,79.5%) from DDRNet repo ,i think results are acceptable,because some basic training tricks such as OHEM are not used.DDRNet-39 gains 79.46% mIoU on Cityscapes val set.

Indeed a great job! You successfully transferred pretrained weights and launched DDRNet training in mmseg. ;)

codecov · 2022-06-29T09:28:18Z

Codecov Report

Attention: Patch coverage is 19.04762% with 136 lines in your changes missing coverage. Please review.

Project coverage is 88.86%. Comparing base (5c113d9) to head (6a2bcb3).
Report is 83 commits behind head on master.

Files	Patch %	Lines
mmseg/models/backbones/ddrnet.py	18.56%	136 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1722      +/-   ##
==========================================
- Coverage   90.24%   88.86%   -1.39%     
==========================================
  Files         142      143       +1     
  Lines        8486     8654     +168     
  Branches     1432     1453      +21     
==========================================
+ Hits         7658     7690      +32     
- Misses        588      724     +136     
  Partials      240      240

Flag	Coverage Δ
unittests	`88.86% <19.04%> (-1.39%)`	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

mmseg/models/backbones/ddrnet.py

MengzhangLI · 2022-06-29T09:31:00Z

Hi, @qyyyyq

Refactory code is necessary and we may leave out more comments in the next few days. You can refer to my re-implementing code of STDC for example.

Original: https://github.com/MichaelFan01/STDC-Seg/blob/master/nets/stdcnet.py
Ours:https://github.com/open-mmlab/mmsegmentation/blob/master/mmseg/models/backbones/stdc.py

Best,

MengzhangLI · 2022-06-29T14:46:10Z

Hi, @qyyyyq
Can you send your wechat id to my e-mail? We could use it to discuss more effectively if you can speak chinese.

tommiekerssies · 2022-10-31T15:06:14Z

What is the status of this PR?

OpenMMLab-Assistant-004 · 2023-04-12T04:31:33Z

Hi @qyyyyq ！We are grateful for your efforts in helping improve mmsegmentation 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❤

Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers. ## Motivation Support DDRNet Paper: [Deep Dual-resolution Networks for Real-time and Accurate Semantic Segmentation of Road Scenes](https://arxiv.org/pdf/2101.06085) official Code: https://github.com/ydhongHIT/DDRNet There is already a PR #1722 , but it has been inactive for a long time. ## Current Result ### Cityscapes #### inference with converted official weights | Method | Backbone | mIoU(official) | mIoU(converted weight) | | ------ | ------------- | -------------- | ---------------------- | | DDRNet | DDRNet23-slim | 77.8 | 77.84 | | DDRNet | DDRNet23 | 79.5 | 79.53 | #### training with converted pretrained backbone | Method | Backbone | Crop Size | Lr schd | Inf time(fps) | Device | mIoU | mIoU(ms+flip) | config | download | | ------ | ------------- | --------- | ------- | ------- | -------- | ----- | ------------- | ------------ | ------------ | | DDRNet | DDRNet23-slim | 1024x1024 | 120000 | 85.85 | RTX 8000 | 77.85 | 79.80 | [config](https://github.com/whu-pzhang/mmsegmentation/blob/ddrnet/configs/ddrnet/ddrnet_23-slim_in1k-pre_2xb6-120k_cityscapes-1024x1024.py) | model \| log | | DDRNet | DDRNet23 | 1024x1024 | 120000 | 33.41 | RTX 8000 | 79.53 | 80.98 | [config](https://github.com/whu-pzhang/mmsegmentation/blob/ddrnet/configs/ddrnet/ddrnet_23_in1k-pre_2xb6-120k_cityscapes-1024x1024.py) | model \| log | The converted pretrained backbone weights download link: 1. [ddrnet23s_in1k_mmseg.pth](https://drive.google.com/file/d/1Ni4F1PMGGjuld-1S9fzDTmneLfpMuPTG/view?usp=sharing) 2. [ddrnet23_in1k_mmseg.pth](https://drive.google.com/file/d/11rsijC1xOWB6B0LgNQkAG-W6e1OdbCyJ/view?usp=sharing) ## To do - [x] support inference with converted official weights - [x] support training on cityscapes dataset --------- Co-authored-by: xiexinch <xiexinch@outlook.com>

…b#1722)

Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers. ## Motivation Support DDRNet Paper: [Deep Dual-resolution Networks for Real-time and Accurate Semantic Segmentation of Road Scenes](https://arxiv.org/pdf/2101.06085) official Code: https://github.com/ydhongHIT/DDRNet There is already a PR open-mmlab#1722 , but it has been inactive for a long time. ## Current Result ### Cityscapes #### inference with converted official weights | Method | Backbone | mIoU(official) | mIoU(converted weight) | | ------ | ------------- | -------------- | ---------------------- | | DDRNet | DDRNet23-slim | 77.8 | 77.84 | | DDRNet | DDRNet23 | 79.5 | 79.53 | #### training with converted pretrained backbone | Method | Backbone | Crop Size | Lr schd | Inf time(fps) | Device | mIoU | mIoU(ms+flip) | config | download | | ------ | ------------- | --------- | ------- | ------- | -------- | ----- | ------------- | ------------ | ------------ | | DDRNet | DDRNet23-slim | 1024x1024 | 120000 | 85.85 | RTX 8000 | 77.85 | 79.80 | [config](https://github.com/whu-pzhang/mmsegmentation/blob/ddrnet/configs/ddrnet/ddrnet_23-slim_in1k-pre_2xb6-120k_cityscapes-1024x1024.py) | model \| log | | DDRNet | DDRNet23 | 1024x1024 | 120000 | 33.41 | RTX 8000 | 79.53 | 80.98 | [config](https://github.com/whu-pzhang/mmsegmentation/blob/ddrnet/configs/ddrnet/ddrnet_23_in1k-pre_2xb6-120k_cityscapes-1024x1024.py) | model \| log | The converted pretrained backbone weights download link: 1. [ddrnet23s_in1k_mmseg.pth](https://drive.google.com/file/d/1Ni4F1PMGGjuld-1S9fzDTmneLfpMuPTG/view?usp=sharing) 2. [ddrnet23_in1k_mmseg.pth](https://drive.google.com/file/d/11rsijC1xOWB6B0LgNQkAG-W6e1OdbCyJ/view?usp=sharing) ## To do - [x] support inference with converted official weights - [x] support training on cityscapes dataset --------- Co-authored-by: xiexinch <xiexinch@outlook.com>

add DDRNet

4ef0a12

mm-assistant bot assigned MeowZheng Jun 29, 2022

MengzhangLI assigned MengzhangLI and unassigned MeowZheng Jun 29, 2022

MengzhangLI reviewed Jun 29, 2022

View reviewed changes

mmseg/models/backbones/ddrnet.py Outdated Show resolved Hide resolved

MengzhangLI changed the title ~~add DDRNet~~ [Feature] Supprt DDRNet Jun 29, 2022

MengzhangLI changed the title ~~[Feature] Supprt DDRNet~~ [Feature] Support DDRNet Jun 29, 2022

change mmseg fomat

6a2bcb3

whu-pzhang mentioned this pull request Apr 6, 2023

[Feature] Support DDRNet #2855

Merged

2 tasks

wjkim81 pushed a commit to wjkim81/mmsegmentation that referenced this pull request Dec 3, 2023

[Feature] Support evaluating CocoMetric without ann_file. (open-mmla…

e927015

…b#1722)

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Feature] Support DDRNet #1722

[Feature] Support DDRNet #1722

qyyyyq commented Jun 29, 2022

CLAassistant commented Jun 29, 2022 •

edited

MengzhangLI commented Jun 29, 2022 •

edited

MengzhangLI Jun 29, 2022

qyyyyq Jun 29, 2022

MengzhangLI Jun 29, 2022

codecov bot commented Jun 29, 2022 •

edited

MengzhangLI commented Jun 29, 2022

MengzhangLI commented Jun 29, 2022 •

edited

tommiekerssies commented Oct 31, 2022

OpenMMLab-Assistant-004 commented Apr 12, 2023

[Feature] Support DDRNet #1722

Are you sure you want to change the base?

[Feature] Support DDRNet #1722

Conversation

qyyyyq commented Jun 29, 2022

Motivation

Modification

Note

CLAassistant commented Jun 29, 2022 • edited

MengzhangLI commented Jun 29, 2022 • edited

MengzhangLI Jun 29, 2022

Choose a reason for hiding this comment

qyyyyq Jun 29, 2022

Choose a reason for hiding this comment

MengzhangLI Jun 29, 2022

Choose a reason for hiding this comment

codecov bot commented Jun 29, 2022 • edited

Codecov Report

MengzhangLI commented Jun 29, 2022

MengzhangLI commented Jun 29, 2022 • edited

tommiekerssies commented Oct 31, 2022

OpenMMLab-Assistant-004 commented Apr 12, 2023

CLAassistant commented Jun 29, 2022 •

edited

MengzhangLI commented Jun 29, 2022 •

edited

codecov bot commented Jun 29, 2022 •

edited

MengzhangLI commented Jun 29, 2022 •

edited