Himanshu Gupta
Bradley Hayes
Zachary Sunberg
Infer pedestrian's intention
Predict pedestrian's behavior given its intention
Plan a path
to its goal location
Reactive Controller
Predict and Act Controller
Issues
Issue
Need a method that determines the optimal action for the vehicle at any given time by reasoning over the uncertainty in pedestrian intention estimation?
Need a method that determines the optimal action for the vehicle at any given time by reasoning over the uncertainty in pedestrian intention estimation?
Bai et. al, ICRA 2015
Bai et. al, ICRA 2015
Effective
Roll-out Policy is important
Bai et. al, ICRA 2015
$$\delta_s \in \{\textbf{Increase Speed, Decrease Speed,}$$ $$\textbf{Maintain Speed, Sudden Brake\}}$$
Bai et. al, ICRA 2015
ISSUES?
Bai et. al, ICRA 2015
Same as previous POMDP
Scenario 1
(Open Field)
Scenario 2 (Cafeteria Setting)
Scenario 3
(L shaped lobby)
# possible actions in POMDP Planning: 4
# possible actions in POMDP Planning: 11
# humans = 100
# humans = 200
# humans = 300
# humans = 400
Scenario 1
Evaluation Metric: Travel Time (in s)
Evaluation Metric: #Outperformed
Scenario 2
\(1D-A^*\)
\(2D-FMM\)
\(2D-PRM\)
Limited Space Planner
Extended Space Planner
Himanshu Gupta
himanshu.gupta@colroado.edu
Extended Space POMDP Planning
(AAMAS 2022)
https://github.com/himanshugupta1009/extended_space_navigation_pomdp
Limited Space
Planner
Extended Space
Planners
\(1D\)-\(A^*\)
\(2D\)-\(NHV\)
Evaluation Metric: Travel Time (in s)
Evaluation Metric: #Outperformed
\(1D-A^*\)
\(2D-NHV\)