1 pose_generation.types

pose_generation.types

Type definitions for finite candidate pose generation.

Candidate generation separates the full sampled shell from the compact valid table. The full shell carries invalidity masks, reason codes, strategy provenance, and optional diagnostics; compact valid views are the actions an actor or oracle selector may actually choose. Any dataset writer that trains a finite-action value model must apply the valid mask before argmax, softmax, loss targets, and bootstrap maximization.

Direction sampling happens on \(\mathbb{S}^2\). Orientation encodings such as R6D describe candidate pose rows; accumulated target visibility is a separate actor-visible directional-memory feature, not an orientation representation.

1.1 Classes

Name	Description
SamplingStrategy	Angular sampling strategy for candidate directions on S^2.
ViewDirectionMode	Base orientation family applied after candidate centers are sampled.
CandidatePositionMode	Spatial prior used to turn raw sphere samples into candidate centers.
CollisionBackend	Backend for collision tests.
CandidateGenerationRuntimeContext	Runtime-only context for target-conditioned candidate generation.
CandidateContext	Mutable full-shell state passed between sampling and pruning rules.
CandidateSamplingResult	Immutable result of candidate sampling and rule-based pruning.

# pose_generation.types { #aria_nbv.pose_generation.types }

`pose_generation.types`

Type definitions for finite candidate pose generation.

Candidate generation separates the full sampled shell from the compact valid
table. The full shell carries invalidity masks, reason codes, strategy
provenance, and optional diagnostics; compact valid views are the actions an
actor or oracle selector may actually choose. Any dataset writer that trains a
finite-action value model must apply the valid mask before argmax, softmax,
loss targets, and bootstrap maximization.

Direction sampling happens on $\mathbb{S}^2$. Orientation encodings such as R6D
describe candidate pose rows; accumulated target visibility is a separate
actor-visible directional-memory feature, not an orientation representation.

## Classes

| Name | Description |
| --- | --- |
| [SamplingStrategy](aria_nbv.pose_generation.types.SamplingStrategy.qmd#aria_nbv.pose_generation.types.SamplingStrategy) | Angular sampling strategy for candidate directions on S^2. |
| [ViewDirectionMode](aria_nbv.pose_generation.types.ViewDirectionMode.qmd#aria_nbv.pose_generation.types.ViewDirectionMode) | Base orientation family applied after candidate centers are sampled. |
| [CandidatePositionMode](aria_nbv.pose_generation.types.CandidatePositionMode.qmd#aria_nbv.pose_generation.types.CandidatePositionMode) | Spatial prior used to turn raw sphere samples into candidate centers. |
| [CollisionBackend](aria_nbv.pose_generation.types.CollisionBackend.qmd#aria_nbv.pose_generation.types.CollisionBackend) | Backend for collision tests. |
| [CandidateGenerationRuntimeContext](aria_nbv.pose_generation.types.CandidateGenerationRuntimeContext.qmd#aria_nbv.pose_generation.types.CandidateGenerationRuntimeContext) | Runtime-only context for target-conditioned candidate generation. |
| [CandidateContext](aria_nbv.pose_generation.types.CandidateContext.qmd#aria_nbv.pose_generation.types.CandidateContext) | Mutable full-shell state passed between sampling and pruning rules. |
| [CandidateSamplingResult](aria_nbv.pose_generation.types.CandidateSamplingResult.qmd#aria_nbv.pose_generation.types.CandidateSamplingResult) | Immutable result of candidate sampling and rule-based pruning. |