search

JingZhang617 / UCNet

UC-Net: Uncertainty Inspired RGB-D Saliency Detectionvia Conditional Variational Autoencoders, CVPR2020
178 stars 25 forks source link

k(latent_dim)=3 in the released model #3

Open ShuhanChen opened 4 years ago

ShuhanChen commented 4 years ago

The provided model in: https://drive.google.com/file/d/1nzGLnlmntTGbcaShfQvE6ouyfWJD-pIB/view?usp=sharing, is trained with k=3? Did you upload the wrong version?

JingZhang617 commented 4 years ago

Yes, it’s the right version. We find relatively stable performance with dimension of z in the range [2-10]. We choose dimension of z as 3 in this version to achieve trade off between performance and efficiency. You can vary dimension of z to play with it.

Cheers, Jing

On Thu, 18 Jun 2020 at 8:51 pm, shuhan notifications@github.com wrote:

The provided model in: https://drive.google.com/file/d/1nzGLnlmntTGbcaShfQvE6ouyfWJD-pIB/view?usp=sharing, is trained with k=3? Did you upload the wrong version?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/JingZhang617/UCNet/issues/3, or unsubscribe https://github.com/notifications/unsubscribe-auth/AE6B4F4D7QX6JXAA7JDS2RDRXHWSVANCNFSM4OBOWKIA .

-- Jing Zhang Ph.D. Student College of Engineering and Computer Science, Australian National University. Email: zjnwpu@gmail.com