1 CandidateDepths

rendering.candidate_depth_renderer.CandidateDepths(
    depths,
    depths_valid_mask,
    poses,
    reference_pose,
    candidate_indices,
    camera,
    p3d_cameras,
)

Typed result for candidate depth rendering.

candidate_indices maps rendered compact rows back to the full candidate shell. That link is required when writing target-RRI labels or selected heavy diagnostics into rollouts.zarr.

1.1 Attributes

Name	Description
depths	Tensor[‘N’, ‘H’, ‘W’] with per-candidate depth maps in metres.
depths_valid_mask	Tensor[‘N’, ‘H’, ‘W’] boolean mask indicating valid depth pixels.
poses	PoseTW (cam2world) for the rendered subset.
reference_pose	PoseTW (ref2world) for the reference frame corresponding to candidates.
candidate_indices	Indices (long) into the full candidate array (pre-render filtering).
camera	Camera calibration (and ref2camera extrinsics) for the rendered subset, same order as depths.
p3d_cameras	PyTorch3D PerspectiveCameras used for rendering. Same externals as camera.

1.2 Methods

Name	Description
to_serializable	Serialize this batch into a cache-friendly CPU payload.
from_serializable	Reconstruct one batch from a serialized payload.

1.2.1 to_serializable

rendering.candidate_depth_renderer.CandidateDepths.to_serializable()

Serialize this batch into a cache-friendly CPU payload.

1.2.2 from_serializable

rendering.candidate_depth_renderer.CandidateDepths.from_serializable(
    payload,
    *,
    device,
)

Reconstruct one batch from a serialized payload.

1.2.2.1 Parameters

Name	Type	Description	Default
payload	dict[str, object]	Serialized payload produced by to_serializable.	required
device	torch.device	Destination device for tensors and wrappers.	required

1.2.2.2 Returns

Name	Type	Description
	'CandidateDepths'	Reconstructed candidate-depth batch.