Google Coral USB Accelerator を試す その3

Posenet をためそう

GitHub - google-coral/project-posenet: Human Pose Detection on EdgeTPU

$ git clone https://github.com/google-coral/project-posenet
$ cd project-posenet
$ sh install_requirements.sh

simple_pose.py

$ python3 simple_pose.py

Inference time: 26ms

Pose Score:  0.61855656
 nose                 x=210  y=152  score=1.0
 left eye             x=224  y=138  score=1.0
 right eye            x=199  y=137  score=1.0
 left ear             x=244  y=135  score=1.0
 right ear            x=182  y=129  score=0.8
 left shoulder        x=268  y=168  score=0.8
 right shoulder       x=160  y=173  score=1.0
 left elbow           x=282  y=255  score=0.6
 right elbow          x=154  y=256  score=0.9
 left wrist           x=236  y=333  score=0.8
 right wrist          x=162  y=299  score=0.6
 left hip             x=324  y=179  score=0.2
 right hip            x=203  y=233  score=0.0
 left knee            x=344  y=86   score=0.8
 right knee           x=170  y=294  score=0.0
 left ankle           x=316  y=175  score=0.1
 right ankle          x=166  y=307  score=0.0

Pose Score:  0.5867945
 nose                 x=398  y=145  score=1.0
 left eye             x=416  y=128  score=1.0
 right eye            x=382  y=127  score=1.0
 left ear             x=457  y=110  score=0.9
 right ear            x=370  y=120  score=0.2
 left shoulder        x=491  y=169  score=0.9
 right shoulder       x=362  y=151  score=0.8
 left elbow           x=465  y=290  score=0.9
 right elbow          x=328  y=246  score=0.8
 left wrist           x=340  y=303  score=0.9
 right wrist          x=233  y=331  score=0.5
 left hip             x=525  y=106  score=0.0
 right hip            x=361  y=161  score=0.3
 left knee            x=501  y=90   score=0.6
 right knee           x=350  y=167  score=0.1
 left ankle           x=504  y=178  score=0.0
 right ankle          x=306  y=176  score=0.1

カメラを使用したデモも、ちゃんと動きました。