ChanganVR
commented
1 year ago Hi @1154001808 you can find all the state information here: https://github.com/vita-epfl/CrowdNav/blob/master/crowd_sim/envs/utils/state.py. The state passed to the policy network consists of the robot's own full state and humans' observable states.
Dear author, hello!Due to a failed installation of rvo2, I cannot run this code, but I have some questions about the code that I would like to consult with you. In sarl.py--class ValueNetwork(nn.Module)--def forward(self, state), I want to know what the form of state is. Could you give me an example？I guess the form should be 'tensor([[ [dg, vpref, vx, vy, r, px, py, vx, vy, r1, d1, r1+r ] , [dg, vpref, vx, vy, r, px, py, vx, vy, r2, d2, r2+r ], ......,[dg, vpref, vx, vy, r, px, py, vx, vy, ri, di, ri+r] ]])'. I don't know if my guess is correct, and I am eager to receive your answer.Thank you very much.