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

Testing time is 0.8 seconds per frame using Titan Xp #10

Open
zhihuilics opened this issue May 20, 2018 · 4 comments
Open

Testing time is 0.8 seconds per frame using Titan Xp #10

zhihuilics opened this issue May 20, 2018 · 4 comments

Comments

@zhihuilics
Copy link

zhihuilics commented May 20, 2018

Hi, Thanks for converting the original code to Torch version.
However, I tested the code and found the running on one frame using Titan Xp is about 0.8 seconds. The original paper claimed this method is real-time.

Could you please give me some advice on this?

Thanks.

@zhihuilics zhihuilics changed the title Testing time is 4 seconds per frame using Titan Xp Testing time is 0.8 seconds per frame using Titan Xp Jun 13, 2018
@RichardMrLu
Copy link

How to detect a video? My gpu is 3000M, when I detect a video, it crashed down for out of memory...

@hjinlee88
Copy link

With updated code from PR, I get 30 fps using 1080ti at video size 640x480.

@tirtile
Copy link

tirtile commented Jun 30, 2018

@hjinlee88 Did you just update code with PR? have you changed others? i got 13fps with 1080ti and the same image size.

@hjinlee88
Copy link

@tirtile that's weird because i got 9~10fps with 1050 as well.
I use pytorch 0.3.1, cuda 9.0 and cudnn 7.0.
I almost change nothing, but PR #4 makes significant improvement on test.py.

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

4 participants