Real-time performance isn't so good

Open wdkwyf opened this issue 7 years ago • 2 comments

Hi, I have tried the inference code in NYU dataset. However, I can't achieve "real-time" performance as mentioned in your paper. For batch size=1: frame rate is 12 For batch size=3: frame rate is 17 Both are not good for real time(less than 24). I wonder why the inference speed isn't good. Did you do infernece in TWO Titans? Thank you very much.

Oct 24 '18 08:10 wdkwyf

@wdkwyf were you able to run it on gpu? When I run the test using "python model/hourglass_um_crop_tiny.py" I am prompted to use CPU for tensorflow. Any advice on how to use GPU?

Jan 14 '19 01:01 Abdul-Mukit

@Abdul-Mukit I think you should update the gpu_config.py. You can refer to https://github.com/melonwan/denseReg/issues/6

Jan 17 '19 07:01 wdkwyf