That clears a lot of things.

Thanks a lot.

If I use the X position of the Ankle Joint, this parameter is not a part of observed state but I can still see it in my model. So does it classify as external tracking data?

Yeah, it makes sense.

Just another thing to clarify, if we use any other info to calculate reward which is not a part of state. Then is this approach valid? I feel this destroys the purpose of RL so need your view on this.

Any description of what each component of vector means is appreciated. To understand one has to dig through the opensim documentation which does not seem very intuitive.

Thanks.

Can everyone share what are all the ideas you have used. I was trying to train a vgg16 on the 96*96 patches and classify them according to ant which is nearest to the center or empty space.