search

zhechen / PLARD

Progressive LiDAR Adaptation for Road Detection
MIT License
160 stars 44 forks source link

Pretrained Model for Training #5

Open csBob123 opened 4 years ago

csBob123 commented 4 years ago

Hi,

Thank you very much for your code. Could I know which model is used as the pre-trained model in your training?

The PSP model on CitySpaces or the resnet101 pre-trained on ImageNet.

Many thanks

zhechen commented 4 years ago

It should be pspnet

On Sat, Jul 4, 2020 at 10:15 PM csBob123 notifications@github.com wrote:

Hi,

Thank you very much for your code. Could I know which model is used as the pre-trained model in your training?

The PSP model on CitySpaces or the resnet101 pre-trained on ImageNet.

Many thanks

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/zhechen/PLARD/issues/5, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABQHNI2U56QWL3C6XXZONLTRZ4MPHANCNFSM4OQL7QOA .

csBob123 commented 4 years ago

It should be pspnet … On Sat, Jul 4, 2020 at 10:15 PM csBob123 @.***> wrote: Hi, Thank you very much for your code. Could I know which model is used as the pre-trained model in your training? The PSP model on CitySpaces or the resnet101 pre-trained on ImageNet. Many thanks — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#5>, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABQHNI2U56QWL3C6XXZONLTRZ4MPHANCNFSM4OQL7QOA .

Thank you for your reply. I tried to train pspnet with resnet101 ImageNet weight initialization and I got a maxF 94%. I found in your paper, using the pspnet can get a MaxF~96%. Did you use the pre-trained pspnet weight on other datasets?

So sorry for my disturbance.

zhechen commented 4 years ago

Yes. The paper should mention that.