stuarteiffert/RNN-for-Human-Activity-Recognition-using-2D-Pose-Input

Performance between methods?

leviethung2103 opened this issue · 0 comments

Hello there,

I would like to ask you about the results in your experiments

  • 2D pose and raw RGB images, which one is better ?
  • Combine the 2D and raw RGB images, how about this combination ?

I am just curious and getting to know about the action recognition

Thank you in advance