LIP_JPPNet icon indicating copy to clipboard operation
LIP_JPPNet copied to clipboard

Time

Open b121351 opened this issue 7 years ago • 3 comments

What is the expected runtime for pose and parsing? When I'm testing with multiple images, >4secs is taken for the first followed by 1sec(apprx.) for the remaining images excluding the load time.

b121351 avatar Jun 05 '18 13:06 b121351

Actually, the inference time depends on your input size. It will be slower for the multi-scale inference.

Engineering-Course avatar Jun 08 '18 07:06 Engineering-Course

Yes! I agree. But the image which is taking around 1 sec when placed at non-first position is taking 4-6 secs when placed at first position. The same thing applies to all the images. That's the issue I'm talking about. The time is calculated excluding the load time i.e., after the printing statement: [*] Load Success .

b121351 avatar Jun 11 '18 05:06 b121351

Hi @Engineering-Course, How to load the model separately ? Please help.

anantaarora avatar Dec 05 '20 11:12 anantaarora