Giter VIP home page Giter VIP logo

Comments (12)

ZhengPeng7 avatar ZhengPeng7 commented on June 12, 2024

嗯, 是的.
用了DUTS_class+COCO-SEG两个数据集.
优化时是两个分别计算, loss叠加起来, 一次backward的.

from gconet_plus.

IceHowe avatar IceHowe commented on June 12, 2024

好的,了解了,感谢

from gconet_plus.

IceHowe avatar IceHowe commented on June 12, 2024

请问没有多GPU运行代码是吗?那个sub by id脚本是不是多次运行的,而不是多GPU一起训练的

from gconet_plus.

ZhengPeng7 avatar ZhengPeng7 commented on June 12, 2024

对的, 那个是之前在dgx服务器上, 把一个setting跑多次的, 并不是多卡哈.

from gconet_plus.

YujieMo avatar YujieMo commented on June 12, 2024

嗯, 是的. 用了DUTS_class+COCO-SEG两个数据集. 优化时是两个分别计算, loss叠加起来, 一次backward的.

您好,请问那为什么在论文中 Training Set的描述那里写的是"We follow the GICD [14] to use DUTS class
as our training set to design the experiments",即只使用DUTS作为训练集呢?

from gconet_plus.

ZhengPeng7 avatar ZhengPeng7 commented on June 12, 2024

我们的所有的模块、实验设计都是仅使用DUTS_class的, 只是在得到所有的已经确定下来的一切(模型设计, 训练策略)后, 在最后使用了所有的已有training sets的组合训练了一下而已. 也就是说其实我们的一些超参的设置也并未针对DUTS_class+COCO-SEG, 而依然是DUTS_class.

from gconet_plus.

YujieMo avatar YujieMo commented on June 12, 2024

我们的所有的模块、实验设计都是仅使用DUTS_class的, 只是在得到所有的已经确定下来的一切(模型设计, 训练策略)后, 在最后使用了所有的已有training sets的组合训练了一下而已. 也就是说其实我们的一些超参的设置也并未针对DUTS_class+COCO-SEG, 而依然是DUTS_class.

感谢您的即时回复,也就是说即使我在代码中只使用DUTS_class进行训练,也可以达到论文中理想的效果对吗

from gconet_plus.

ZhengPeng7 avatar ZhengPeng7 commented on June 12, 2024

不同训练集的结果可见于. 例如, 仅用DUTS_class训练就是仅用DUTS_class的performance (CoCA的Emax可达0.786). 还有问题欢迎继续留言哈.

from gconet_plus.

YujieMo avatar YujieMo commented on June 12, 2024

不同训练集的结果可见于. 例如, 仅用DUTS_class训练就是仅用DUTS_class的performance (CoCA的Emax可达0.786). 还有问题欢迎继续留言哈.

好的,十分感谢您的回复

from gconet_plus.

zhi-wang-lab avatar zhi-wang-lab commented on June 12, 2024

你的cocoseg数据格式是怎么处理的,我记得cocoseg数据集还包括一个npy文件

from gconet_plus.

IceHowe avatar IceHowe commented on June 12, 2024

你的cocoseg数据格式是怎么处理的,我记得cocoseg数据集还包括一个npy文件

我印象里没有你所说的文件,好像是直接下载就能用的,你可能下错了?

from gconet_plus.

ZhengPeng7 avatar ZhengPeng7 commented on June 12, 2024

是的, 没有npy的吧, 我的google-drive上有, 你可以下一下看看.

from gconet_plus.

Related Issues (10)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.