Skip to content

Jointly learning policies and latent representations for driver behavior.

Notifications You must be signed in to change notification settings

sisl/latent_driver

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

44 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

latent_driver

Jointly learning policies and latent representations for driver behavior. See paper here.

Note that this work uses a now outdated version of AutomotiveDrivingModels.jl -- an old commit can be found here.

The video below illustrates the different driver classes used in training the encoder and policies.

Below we can see the how the encoder chooses to represent trajectories from different driver classes as training progresses.

Once we have a trained policy, we can propagate trajectories by passing observations and samples from the latent space into the policy and using the actions to propagate the scene forward. If we initialize a vehicle at 20 m/s and an aggressive latent state, we can see that it chooses to accelerate.

Instead, if a vehicle is initialized with a passive latent state, it chooses to decelerate.

About

Jointly learning policies and latent representations for driver behavior.

Resources

Stars

Watchers

Forks

Packages

No packages published