rlmesh.adapters.ImageTag

classfrom rlmesh.adapters import ImageTag[source]

A camera image entry in an environment observation.

ImageTag(
    role: str,
    layout: ImageLayout = 'hwc',
    upside_down: bool = False,
    *,
    part: str | None = None,
)

Attributes

roleSemantic role used for matching, e.g.
layoutAxis layout of the stored image.
upside_downWhether the image is rendered rotated 180 degrees (a true rotation, not a vertical flip) relative to the canonical upright orientation.
partThe body part this camera sits on, when the role repeats across a body ("head", "left_arm", ...).

role

attribute[source]

role: str

Semantic role used for matching, e.g. image/primary.

upside_down

attribute[source]

upside_down: bool

Whether the image is rendered rotated 180 degrees (a true rotation, not a vertical flip) relative to the canonical upright orientation. Declared on both ends; the adapter flips only when the env and the model disagree. (If a second orientation is ever needed, this should become a constrained string like dtype, not a wider bool.)

part

attribute[source]

part: str | None

The body part this camera sits on, when the role repeats across a body ("head", "left_arm", …). An identity key the resolver matches on, never a value it checks. Keyword-only and omitted from the wire when unset.