search

zizhaozhang / unet-tensorflow-keras

A concise code for training and evaluating Unet using tensorflow+keras
MIT License
186 stars 72 forks source link

这个代码可以用来分割多类吗？ #2

Closed Kaido0 closed 6 years ago

zizhaozhang commented 7 years ago

可以。需要将conv10改为输出多个channels（类的个数）。然后用softmax + cross_entropy loss

zhcv commented 6 years ago

我多跑分类不行啊，标签怎么制作呢

zizhaozhang commented 6 years ago

set opt.num_class to the one you want. The label is a HxW map where each pixel has value [0,...,#of class).

zhcv commented 6 years ago

I have tried , it does not work , did you run multi label segment with this repository?

zizhaozhang commented 6 years ago

Probably you can show me the error to let me check. It should work with very minor modifications.

On Wed, Nov 22, 2017 at 9:35 AM, zhapeng notifications@github.com wrote:

I have tried , it does not work , did you run multi label segment with this repository?

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/zizhaozhang/unet-tensorflow-keras/issues/2#issuecomment-346367725, or mute the thread https://github.com/notifications/unsubscribe-auth/AFfrUe775gq3VN7_KdzkC-9WE3zOjoSzks5s5DFMgaJpZM4NRijV .

-- Best Regards, Zizhao