search

felixchenfy / Realtime-Action-Recognition

Apply ML to the skeletons from OpenPose; 9 actions; multiple people. (WARNING: I'm sorry that this is only good for course demo, not for real world applications !!! Those ary very difficult !!!)
MIT License
857 stars 248 forks source link

about openpose model #46

Open kyriefeng opened 4 years ago

kyriefeng commented 4 years ago

how can i choose the model of openpose during training？ the default is mobilenet_thin！ i wanna to compare accuracy using different models

felixchenfy commented 4 years ago

In the config file, see this paragraph:

openpose:
      model: mobilenet_thin
      img_size: 656x368
      # model: cmu
      # img_size: 432x368

You can comment out line 1&2, and uncomment line 3&4.

Currently, there are only these 2 models.

seung2jin commented 4 years ago

hello @felixchenfy such a great project. i wanna change openpose model to openpose-darknet. may i get some advise??

bokey007 commented 4 years ago

hi, for better opse estimation Just change model = cmu