hi there
I started to use the pose tracker which an awesome tool that gives a good opportunity to animate the character for free , however, the quality is a bit bad, such as the shaking body and sliding feet ! I am not an expert but why the extension is limited only to one cam ?
most of motion capture software that I came through suggest to use at least two cams to achieve good results " more cams = a stable character moves " ! like move ai which uses apple devices !
for me, I have 4 cams that belong to android category !
so, is it possible, in the future, to use more than one cam in making motion captures within " pose tracker" ?
Hello @Massinissa! Here is a link to our documentation for the Pose Tracker. Pose tracker uses a fully convolutional neural network model to identify body positions within the frame. You can take a look at the demonstration to insure that the video you are recording the motion from has the best results. If you would like, you could submit a video, and if possible, your USD file, of the issues you are having and I will send it over to the dev team to take a look. You can attach it here or email it to me at omniversecommunity@nvidia-omniverse.atlassian.net
I will submit a request to the dev team to support more cameras!
A feature request ticket was created from this post. OM-84597: Pose Tracker Multiple Camera Support
thanks for the reply , I found this video and it clarified al of things, and it brought a bit of stability by " straps as tracks or green screen " . however, inside machinima, green screen does not fit until you add something so the engine starts to work
here the link to the video :
