Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Pretrained weights of Resnet-34 and Resnet-101 backbone #270

Open
sjiang95 opened this issue Dec 2, 2021 · 8 comments
Open

Pretrained weights of Resnet-34 and Resnet-101 backbone #270

sjiang95 opened this issue Dec 2, 2021 · 8 comments

Comments

@sjiang95
Copy link

sjiang95 commented Dec 2, 2021

(This is NOT an issue)

NOTICE! The performance (inference accuracy, avg FPS, etc.) of the shared weights is untested.

You can download them from OneDrive or BaiduCloud(code:srni).

Table of files:

Trained on Dataset Resnet34 Resnet101
CULane culane_34.pth culane_101.pth
TuSimple tusimple_34.pth tusimple_101.pth

Please refer to this table for integrity check.

Filename SHA256
culane_34.pth e3acebccc1a320453c4504257a41b64c361a6c2109d429559cb9ddebc28d9bb9
culane_101.pth ab1f227a70a6936cdd343204eb83a99c3d76d4c11d12192afac5ae0f0d51c7f8
tusimple_34.pth 7e81c26d5268a28c8c1ee72b4878c2f287420989f1cfc39379b8ce2b43231c0b
tusimple_101.pth dff176f951fa2cacfa268248f3e18db8059a9f3c2a6a2382459368376c171300
@sjiang95 sjiang95 closed this as completed Dec 2, 2021
@sjiang95 sjiang95 reopened this Dec 2, 2021
@cfzd
Copy link
Owner

cfzd commented Dec 3, 2021

@AdorableJiang
Thanks for your contribution! 👍

@xqh5201314
Copy link

你好,我使用了您提供的resnet34和101的模型去测试我的视频,发现结果里面什么也没有检测到,请问是什么原因呢?

@sjiang95
Copy link
Author

sjiang95 commented Jun 5, 2022

你好,我使用了您提供的resnet34和101的模型去测试我的视频,发现结果里面什么也没有检测到,请问是什么原因呢?

您提供的信息不足,我只能猜测configs/culane.pyconfigs/tusimple.py 中的backbone = 一项有改到和使用的backbone相匹配吗?

@scyyh11
Copy link

scyyh11 commented Aug 18, 2023

你好,请问这个权重经过了多少轮的预训练?谢谢

@sjiang95
Copy link
Author

你好,请问这个权重经过了多少轮的预训练?谢谢

您好,由于年代久远,我只能猜测当时应该使用了 https://github.com/cfzd/Ultra-Fast-Lane-Detection/tree/353df107756b8c03c22c27201e33fc63d84ecfe6/configs 中的默认训练设置。

@scyyh11
Copy link

scyyh11 commented Aug 18, 2023

你好,请问这个权重经过了多少轮的预训练?谢谢

您好,由于年代久远,我只能猜测当时应该使用了 https://github.com/cfzd/Ultra-Fast-Lane-Detection/tree/353df107756b8c03c22c27201e33fc63d84ecfe6/configs 中的默认训练设置。

谢谢您及时的回复,使用res34进行验证后发现效果与论文内相似,应该是默认训练设置中的50轮

@zswl0117
Copy link

@sjiang95 您好,非常感谢您提供权重代码
我测试了您的代码如第一张图,但是发现和原文的给的数据(如第二张图)还是有较大出入?
请问是不是我的测试方法有问题
{BE751D67-D78D-4c9c-A3C9-AEB98F3A3F7D}
image

@sjiang95
Copy link
Author

@sjiang95 您好,非常感谢您提供权重代码 我测试了您的代码如第一张图,但是发现和原文的给的数据(如第二张图)还是有较大出入? 请问是不是我的测试方法有问题 {BE751D67-D78D-4c9c-A3C9-AEB98F3A3F7D} image

您好,我跟您一样尝试在自己的项目中部署该文章的模型,开这个issue只是共享训练好的权重以方便有需要的人。
我并非该项目作者,所以无法回答您提出的与原文结果有出入的问题。

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

5 participants