Ultra-Fast-Lane-Detection-V2

PyTorch实现的论文“https://arxiv.org/abs/2206.07389”。

PyTorch implementation of the paper "https://arxiv.org/abs/2206.07389".

! [] (ufldv2.png“能见度”)

#演示

Demo

< a href = " https://youtu。

<a href="https://youtu.

be/VkvpoHlaMe0" target="_blank">

#安装

Install

请参见INSTALL.md

Please see INSTALL.md

#开始行动

Get started

请在您想要运行的任何配置中修改' data_root '。

Please modify the data_root in any configs you would like to run.

我们将使用' configs/culane_res18.py '作为示例。

We will use configs/culane_res18.py as an example.

要训练模型，可以运行:

To train the model, you can run:

Python train.py configs/culane_res18.py——log_path /path/to/your/work/dir

python train.py configs/culane_res18.py --log_path /path/to/your/work/dir

或

or

Python -m torch. distribudit .launch——nproc_per_node=8 train.py configs/culane_res18.py——log_path /path/to/your/work/dir

python -m torch.distributed.launch --nproc_per_node=8 train.py configs/culane_res18.py --log_path /path/to/your/work/dir

需要注意的是，如果使用不同数量的gpu，学习率也要相应调整。

It should be noted that if you use different number of GPUs, the learning rate should be adjusted accordingly.

配置的学习率对应于CULane和CurveLanes数据集上的8个gpu训练。

The configs' learning rates correspond to 8-GPU training on CULane and CurveLanes datasets.

*如果你想用单个GPU在CULane或curvelane上进行训练，请将学习率降低1/8。

*If you want to train on CULane or CurveLanes with single GPU, please decrease the learning rate by a factor of 1/8.

*在tussimple上，学习率对应于单个GPU的训练。

On the Tusimple, the learning rate corresponds to single GPU training.

#训练有素的模型

Trained models

我们提供CULane, Tusimple和CurveLanes上的训练模型。

We provide trained models on CULane, Tusimple, and CurveLanes.

开始使用

“data_root”。

请修改你想要运行的任何配置文件中的 data_root。

我们将以配置/ culane_res18.py为例。

我们将以 configs/culane_res18.py 为例。

要训练模型，你可以运行：

Python train.py configs/culane_res18.py——log_path /path/to/your/work/dir

python train.py configs/culane_res18.py --log_path /path/to/your/work/dir

python G:/postgraduate_studyfile/Ultra-Fast-Lane-Detection-v2-master/train.py configs/tusimple_res18.py——log_path G:/postgraduate_studyfile/Ultra-Fast-Lane-Detection-v2-master/model_data .py

python G:/postgraduate_studyfile/Ultra-Fast-Lane-Detection-v2-master/train.py configs/tusimple_res18.py --log_path G:/postgraduate_studyfile/Ultra-Fast-Lane-Detection-v2-master/model_data

或者

Python -m torch. distribudit .launch——nproc_per_node=8 train.py configs/culane_res18.py——log_path /path/to/your/work/dir

python -m torch.distributed.launch --nproc_per_node=8 train.py configs/culane_res18.py --log_path /path/to/your/work/dir

(2)图形处理器，图形处理器，图形处理器，图形处理器。

需要注意的是，如果你使用不同数量的GPU，学习率应该相应地调整。

8 .图形处理器，图形处理器

配置文件中的学习率对应于在CULane和CurveLanes数据集上使用8个GPU进行训练。

1/8。

*如果你想在CULane或CurveLanes上使用单个GPU进行训练，请将学习率减少1/8倍。

*在Tusimple上，学习率对应于单个GPU训练。

训练好的模型

曲线图，曲线图，曲线图。

我们提供了在CULane，Tusimple和CurveLanes上训练好的模型。

|------------|----------|-------|------|

|图森| ResNet18 | 96.11 | https://drive.google.com/file/d/1Clnj9-dLz81S3wXiYtlkc4HVusCb978t/view?usp=sharing/https://pan.baidu.com/s/1umHo0RZIAQ1l_FzL2aZomw?pwd=6xs1 |

|图森| ResNet34 | 96.24 | https://drive.google.com/file/d/1pkz8homK433z39uStGK3ZWkDXrnBAMmX/view?usp=sharing/https://pan.baidu.com/s/1Eq7oxnDoE0vcQGzs1VsGZQ?pwd=b88p |

要进行评估，请运行

For evaluation, run

壳牌

Shell

mkdir tmp

Python test.py configs/culane_res18.py——test_model /path/to/your/model.pth——test_work_dir ./tmp

python test.py configs/culane_res18.py --test_model /path/to/your/model.pth --test_work_dir ./tmp

与训练相同，还支持多gpu评估。

Same as training, multi-gpu evaluation is also supported.

壳牌

Shell

mkdir tmp

——test_model /path/to/your/model.pth——test_work_dir ./tmp

python -m torch.distributed.launch --nproc_per_node=8 test.py configs/culane_res18.py --test_model /path/to/your/model.pth --test_work_dir ./tmp

#可视化

Visualization

我们提供了一个脚本来可视化检测结果。

We provide a script to visualize the detection results.

运行以下命令以可视化CULane的测试集。

Run the following commands to visualize on the testing set of CULane.

Python demo.py configs/culane_res18.py——test_model /path/to/your/culane_res18.pth

python demo.py configs/culane_res18.py --test_model /path/to/your/culane_res18.pth

tensort部署

Tensorrt Deploy

我们还提供了一个python脚本来对视频进行张排序推理。

We also provide a python script to do tensorrt inference on videos.

转换为onnx模型

Convert to onnx model

Python deploy/pt2onnx.py——config_path configs/culane_res34.py——model_path weights/culane_res34.pth

python deploy/pt2onnx.py --config_path configs/culane_res34.py --model_path weights/culane_res34.pth

也可以通过以下脚本下载onnx模型:https://github.com/PINTO0309/PINTO_model_zoo/blob/main/324_Ultra-Fast-Lane-Detection-v2/download.sh。

Or you can download the onnx model using the following script: https://github.com/PINTO0309/PINTO_model_zoo/blob/main/324_Ultra-Fast-Lane-Detection-v2/download.sh.

并复制' ufldv2_culane_res34_320x1600。

And copy `ufldv2_culane_res34_320x1600.

到' weights/ufldv2_culane_res34_320x1600.onnx '

onnxtoweights/ufldv2_culane_res34_320x1600.onnx`

转换为张排序模型

Convert to tensorrt model

使用trtexec转换引擎模型

Use trtexec to convert engine model

“trtexec——onnx =重量/ culane_res34。

`trtexec --onnx=weights/culane_res34.

onnx——saveEngine =重量/ culane_res34.engine '

onnx --saveEngine=weights/culane_res34.engine`

做推理

Do inference

——config_path configs/culane_res34.py——engine_path weights/culane_res34. py

python deploy/trt_infer.py --config_path configs/culane_res34.py --engine_path weights/culane_res34.

引擎——video_path example.mp4

engine --video_path example.mp4

#引用

Citation

助理

BibTeX

@InProceedings {qin2020ultra,

@InProceedings{qin2020ultra,

作者={秦，泽群，王，环宇，李，喜}，

author = {Qin, Zequn and Wang, Huanyu and Li, Xi},

title ={超快速结构感知深车道检测}，

title = {Ultra Fast Structure-aware Deep Lane Detection},

{欧洲计算机视觉会议(ECCV)}，

booktitle = {The European Conference on Computer Vision (ECCV)},

年份= {2020}

year = {2020}

}

@ARTICLE {qin2022ultrav2,

@ARTICLE{qin2022ultrav2,

作者={秦，泽群，张，彭义，李，喜}，

author={Qin, Zequn and Zhang, Pengyi and Li, Xi},

{IEEE模式分析与机器智能汇刊}，

journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},

title={基于混合锚驱动有序分类的超快速深巷检测};

title={Ultra Fast Deep Lane Detection With Hybrid Anchor Driven Ordinal Classification},

年= {2022},

year={2022},

体积= {},

volume={},

数量= {},

number={},

页面= {1 - 14},

pages={1-14},

doi = {10.1109 / TPAMI.2022。

doi={10.1109/TPAMI.2022

Ultra-Fast-Lane-Detection-V2

PyTorch implementation of the paper "Ultra Fast Deep Lane Detection with Hybrid Anchor Driven Ordinal Classification".

Demo

Install

Please see INSTALL.md

Get started

Please modify the data_root in any configs you would like to run. We will use configs/culane_res18.py as an example.

To train the model, you can run:

python train.py configs/culane_res18.py --log_path /path/to/your/work/dir

or

python -m torch.distributed.launch --nproc_per_node=8 train.py configs/culane_res18.py --log_path /path/to/your/work/dir

It should be noted that if you use different number of GPUs, the learning rate should be adjusted accordingly. The configs' learning rates correspond to 8-GPU training on CULane and CurveLanes datasets. If you want to train on CULane or CurveLanes with single GPU, please decrease the learning rate by a factor of 1/8. On the Tusimple, the learning rate corresponds to single GPU training.

Trained models

We provide trained models on CULane, Tusimple, and CurveLanes.

开始使用请修改你想要运行的任何配置文件中的 data_root。我们将以 configs/culane_res18.py 为例。

要训练模型，你可以运行：

python train.py configs/culane_res18.py --log_path /path/to/your/work/dir python G:/postgraduate_studyfile/Ultra-Fast-Lane-Detection-v2-master/train.py configs/tusimple_res18.py --log_path G:/postgraduate_studyfile/Ultra-Fast-Lane-Detection-v2-master/model_data

或者

python -m torch.distributed.launch --nproc_per_node=8 train.py configs/culane_res18.py --log_path /path/to/your/work/dir

需要注意的是，如果你使用不同数量的GPU，学习率应该相应地调整。配置文件中的学习率对应于在CULane和CurveLanes数据集上使用8个GPU进行训练。*如果你想在CULane或CurveLanes上使用单个GPU进行训练，请将学习率减少1/8倍。*在Tusimple上，学习率对应于单个GPU训练。

训练好的模型我们提供了在CULane，Tusimple和CurveLanes上训练好的模型。

Dataset	Backbone	F1	Link
CULane	ResNet18	75.0	Google/Baidu
CULane	ResNet34	76.0	Google/Baidu
Tusimple	ResNet18	96.11	Google/Baidu
Tusimple	ResNet34	96.24	Google/Baidu
CurveLanes	ResNet18	80.42	Google/Baidu
CurveLanes	ResNet34	81.34	Google/Baidu

For evaluation, run

mkdir tmp

python test.py configs/culane_res18.py --test_model /path/to/your/model.pth --test_work_dir ./tmp

Same as training, multi-gpu evaluation is also supported.

mkdir tmp

python -m torch.distributed.launch --nproc_per_node=8 test.py configs/culane_res18.py --test_model /path/to/your/model.pth --test_work_dir ./tmp

Visualization

We provide a script to visualize the detection results. Run the following commands to visualize on the testing set of CULane.

python demo.py configs/culane_res18.py --test_model /path/to/your/culane_res18.pth

Tensorrt Deploy

We also provide a python script to do tensorrt inference on videos.

Convert to onnx model
```
python deploy/pt2onnx.py --config_path configs/culane_res34.py --model_path weights/culane_res34.pth
```
Or you can download the onnx model using the following script: https://github.com/PINTO0309/PINTO_model_zoo/blob/main/324_Ultra-Fast-Lane-Detection-v2/download.sh. And copy ufldv2_culane_res34_320x1600.onnx to weights/ufldv2_culane_res34_320x1600.onnx
Convert to tensorrt model

Use trtexec to convert engine model

trtexec --onnx=weights/culane_res34.onnx --saveEngine=weights/culane_res34.engine

Do inference

python deploy/trt_infer.py --config_path  configs/culane_res34.py --engine_path weights/culane_res34.engine --video_path example.mp4

Citation

@InProceedings{qin2020ultra,
author = {Qin, Zequn and Wang, Huanyu and Li, Xi},
title = {Ultra Fast Structure-aware Deep Lane Detection},
booktitle = {The European Conference on Computer Vision (ECCV)},
year = {2020}
}

@ARTICLE{qin2022ultrav2,
  author={Qin, Zequn and Zhang, Pengyi and Li, Xi},
  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence}, 
  title={Ultra Fast Deep Lane Detection With Hybrid Anchor Driven Ordinal Classification}, 
  year={2022},
  volume={},
  number={},
  pages={1-14},
  doi={10.1109/TPAMI.2022.3182097}
}

Name		Name	Last commit message	Last commit date
Latest commit History 93 Commits
configs		configs
data		data
deploy		deploy
evaluation		evaluation
model		model
my_interp		my_interp
scripts		scripts
utils		utils
zl_file		zl_file
.gitignore		.gitignore
INSTALL.md		INSTALL.md
LICENSE		LICENSE
README.md		README.md
UFLD2在WSL的使用.txt.lnk		UFLD2在WSL的使用.txt.lnk
Ultra-Fast-Lane-Detection-v2-master.code-workspace		Ultra-Fast-Lane-Detection-v2-master.code-workspace
conda_env.txt		conda_env.txt
demo.py		demo.py
example.mp4		example.mp4
import.txt		import.txt
requirement.txt		requirement.txt
requirements.txt		requirements.txt
speed_simple.py		speed_simple.py
strach.ipynb		strach.ipynb
test.py		test.py
train.py		train.py

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

Ultra-Fast-Lane-Detection-V2

Ultra-Fast-Lane-Detection-V2

Demo

Install

Get started

Trained models

Visualization

tensort部署

Tensorrt Deploy

Citation

Ultra-Fast-Lane-Detection-V2

Demo

Install

Get started

Trained models

Visualization

Tensorrt Deploy

Citation

About

Releases

Packages

Languages

License

FriendChi/dis-dark-res34

Folders and files

Latest commit

History

Repository files navigation

Ultra-Fast-Lane-Detection-V2

Ultra-Fast-Lane-Detection-V2

Demo

Install

Get started

Trained models

Visualization

tensort部署

Tensorrt Deploy

Citation

Ultra-Fast-Lane-Detection-V2

Demo

Install

Get started

Trained models

Visualization

Tensorrt Deploy

Citation

About

Resources

License

Stars

Watchers

Forks

Releases

Packages 0

Languages

Packages