Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Display attention lines #119

Open
IsYang23 opened this issue May 31, 2024 · 1 comment
Open

Display attention lines #119

IsYang23 opened this issue May 31, 2024 · 1 comment

Comments

@IsYang23
Copy link

  1. Hey sir, Thanks for your great work and contributions. I have run the script for highway_env with DQN with ego_attention. But I cannot figure out how to display the attention lines (like you in your int_ego_attention experiment). When you record the videos, is there a way to display the attention lines? If not, what other ways can i use to observe the attention of the vehicle? I am interested in ways to visualize them. Maybe attention maps? Or heat maps? But as a beginner in the field I don't know how to figure that out yet, I was hoping you could help me.

  2. Another thing is about the vehicles velocity, I want to plot the velocity vs time for the entire training duration, I don't know how to do it.

  3. Also Tensorboard's plots don't have axis names on the plots, is there a way to modify the plots displayed by tensorboard (logs)? And/or add some new plots to it from ?

https://github.com/eleurent/rl-agents/assets/149097708/e0016937-65e5-4724-9cae-ccc2a5dd0d0b
4. In this video, the agent keep going back and forth from one lane to another, I am thinking this behavior is linked to the reward of lane changing? What do you think? How can it be adjusted to avoid it happening in this case?

Sorry if it is a bit messy, please feel free to answer at your convenience.

@IsYang23
Copy link
Author

I saw a recent closed issue on vehicle trajectory display, I will try to implement it. Thanks a lot already!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant