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

精度提升 #18

Open
Thea1 opened this issue Sep 17, 2020 · 4 comments
Open

精度提升 #18

Thea1 opened this issue Sep 17, 2020 · 4 comments

Comments

@Thea1
Copy link

Thea1 commented Sep 17, 2020

我比较早看到你的论文和代码,觉得写的很好,我也跑到了和之前论文中差不多的精度,但是最近发现,你的baseline精度又提升了,但是看代码好像并无太大变化,就想请教一下,您是如何在aircraft数据集上将精度从94.50%提升到94.70%,以及在car数据集上从94.70%提升到95.00%?

@ZF4444
Copy link
Owner

ZF4444 commented Sep 17, 2020

@Thea1
Copy link
Author

Thea1 commented Sep 17, 2020

目前我跑的精度分别为,cub:89.25%,aircraft : 94.06%,cars:94.75%。
想知道如何调整才能达到你论文的精度呢? 期待您的回复,谢谢!

@Thea1
Copy link
Author

Thea1 commented Sep 17, 2020

谢谢!主要是augmentation添加了color distort
https://github.com/ZF1044404254/MMAL-Net/blob/199c1706695ba0f645809d3568881816fbe2fb70/datasets/dataset.py#L62

哇喔 回的好快 受宠若惊 知道啦 我去改一改试试 非常感谢!!!

@ZF4444
Copy link
Owner

ZF4444 commented Sep 17, 2020

目前我跑的精度分别为,cub:89.25%,aircraft : 94.06%,cars:94.75%。
想知道如何调整才能达到你论文的精度呢? 期待您的回复,谢谢!

我是对100epoch 后每个epoch都进行test,得到最高的精度。其次由于没有作随机数等的限制,每次训练结果都会有些许差异,或许多跑几次。

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

2 participants