vizdoom_nav
---------------------

This folder contains videos of image based navigation trials in the VizDoom simulation environment. 


# Explanations for the videos:
 
- The top 4 images on the left of the videos show the current-state seen by the agent. 
Images correspond to forward/right/backward/left directions, and have a resolution of 160 x 120.

- The bottom 4 static images on the left of the videos show the goal-state the agent is trying to navigate to. 
Images correspond to forward/right/backward/left directions, and have a resolution of 160 x 120.

- The layout on the right of the videos provides a top-down visualization of the navigation trial.
We emphasize that the agent only receives current and goal images as input, and the top-down layout 
is solely for visualization purposes.


# Additional explanations for the top-down layout visualization:

- The black dots mark the positions of column like obstructions in the map (e.g., trees, torches, lamps)

- The white box with the cross sign marks the position of the goal-state.

- The moving black box marks the current position of the agent, the yellow box with a diagonal
line marks the starting position, and the smaller yellow boxes in-between mark the past positions
visited by the agent.

- The blue trajectories that blip in and out of view visualize the restitched trajectories resulting
from retrospective planning. As the agent replans at each timestep in a model predictive control loop,
there is a different restitched trajectory in each frame.

