Skip to content

Latest commit

 

History

History
85 lines (49 loc) · 3.01 KB

replicant_dynamic.md

File metadata and controls

85 lines (49 loc) · 3.01 KB

ReplicantDynamic

from tdw.replicant.replicant_dynamic import ReplicantDynamic

Dynamic data for a Replicant that can change per communicate() call (such as the position of the Replicant).

Fields

  • held_objects A dictionary of objects held in each hand. Key = Arm. Value = Object ID.

  • body_parts Transform data for each body part. Key = Body part ID. Value = Transform.

  • collisions Collision data per body part. Key = Body part ID. Value = A list of object IDs that the body part collided with.

  • output_data_status This is meant for internal use only. For certain actions, the build will update the Replicant's ActionStatus. Do not use this field to check the Replicant's status. Always check replicant.action.status instead.

  • transform The Transform of the agent.

  • images The images rendered by the agent as dictionary. Key = the name of the pass. Value = the pass as a numpy array.

  • projection_matrix The camera projection matrix of the agent's camera as a numpy array.

  • camera_matrix The camera matrix of the agent's camera as a numpy array.

  • got_images If True, we got images from the output data.

  • avatar_id The ID of the avatar.

Functions

__init__

ReplicantDynamic(resp, replicant_id, frame_count)

Parameter Type Default Description
resp List[bytes] The response from the build, which we assume contains replicant output data.
replicant_id int The ID of this replicant.
frame_count int The current frame count.

save_images

self.save_images(output_directory)

Save the ID pass (segmentation colors) and the depth pass to disk. Images will be named: [frame_number]_[pass_name].[extension] For example, the depth pass on the first frame will be named: 00000000_depth.png

The img pass is either a .jpg. The id and depth passes are .png files.

Parameter Type Default Description
output_directory PATH The directory that the images will be saved to.

get_pil_image

self.get_pil_image()

self.get_pil_image(pass_mask="img")

Convert raw image data to a PIL image. Use this function to read and analyze an image in memory. Do NOT use this function to save image data to disk; save_image is much faster.

Parameter Type Default Description
pass_mask str "img" The pass mask. Options: "img", "id", "depth".

Returns: A PIL image.

get_collision_enters

self.get_collision_enters(collision_detection)

Parameter Type Default Description
collision_detection CollisionDetection The CollisionDetection rules.

Returns: A list of body IDs that entered a collision on this frame, filtered by the collision detection rules.