dlunion / easyopenpose Goto Github PK

Hi! Thank you for your project! I have a doubt about the none-maximum suppression in:
https://github.com/dlunion/EasyOpenPose/blob/master/easyOpenPose.cpp#L449
I don't understand why computing a weighted coordinates (weighted by its confidence?) in a region of 7*7.
Looking forward to your help. Thank you!

hi, I am very sorry to trouble you. I have some question to implement the function.

1. Download the pose_iter_440000.caffemodel,Where do I put the model?
   2)How I will execute the generated file? How to pass the reference?
   3)you file EasyOpenPose Whether it should be written OpenPose?
   I am looking forward to your reply?
   Thank you.

您好，该模型支持手部关键点检测吗？

This project is awesome!

BTW, where did u find the pre-trained model?

I searched the original OpenPose project, but found nothing about the pre-trained model or the training codes?

Thank u :)

Sorry for disturbing u, I find myself. :P

hi,Thank you very much for your demo. I am very sorry to trouble you. I want to ask you a question
Usage rate of GPU ,the number of GPU ,the type of your GPU,and Usage rate of cpu . Thank you for your reply! Thank you!

Good Job! Simple and reliable!

examples/EasyOpenPose/easyOpenPose.cpp:726:16: error: no matching function for call to ‘caffe::Net<float>::Forward()’
  net_->Forward();
                ^
In file included from ./include/caffe/caffe.hpp:12:0,
                 from examples/EasyOpenPose/easyOpenPose.cpp:2:
./include/caffe/net.hpp:53:31: note: candidate: const std::vector<caffe::Blob<Dtype>*>& caffe::Net<Dtype>::Forward(const std::vector<caffe::Blob<Dtype>*>&, Dtype*) [with Dtype = float]
   const vector<Blob<Dtype>*>& Forward(const vector<Blob<Dtype>* > & bottom,

some help me! thanks

image: test.jpg
deploy: pose_deploy_linevec.prototxt
caffemodel: pose_iter_440000.caffemodel
compute device GPU(0).
base size = width 500 x height 500
forward fee: 9989.028 ms
finish. save result to 'test_openpose.jpg', people: 3