Soft-NMS

Soft-NMS copied to clipboard

gpu softnms is slower because you use numpy in the middle of a torch ops

4

torch execute gpu ops in async, so when you call numpy, torch will sync up all ops and transfer to cpu, then transfer back when you call torch ops again,...

zylo117

if tscore < maxscore:

1

请问下边的代码是什么意思呢？ sc不是按大到小排序的，这行代码不会执行呀？

yja1

Refactors include improved readibility in softnms_pytorch.py and consistent use of pytorch to improve performance.

imadtoubal

I have used this code to evaluate my own project SSD, but it seems that soft-nms-gaussian doesn't use the nms thread , it only have two parameters(sigma, thresh) , and...

lucky-xu-1994

Is there an offical tensorflow version of soft nms?

1

Hi, Thanks for shaing your code. I want to ask is there an offical tensorflow version of soft nms. If I use a self-written cpu version of soft nms, like...

zpge

juvekaradheesh