AutoMovie API
    Preparing search index...

    Namespace HumanHoldMesh

    A two-person hold's key pose adapted to any pair of bodies (#3438): the joints of both bodies at the key pose, the target bodies' bone lengths and the hold's contacts go through the interaction mesh (HumanInteractionMesh, Ho, Komura and Tai 2010), and each body's adapted joints become its limb goals (each wrist and ankle on its adapted joint, bent toward its adapted elbow or knee) and its root's shift, which the existing limb fits (resolveHumanPosture) resolve.

    The key pose is a coupled solve of the hold on two reference bodies (#3437), never authored per hold; a contact keeps the offset between its two joints that the key pose has, so a hand on a back stays a hand's thickness off the spine joint.

    First pass: a contact's offset is kept as the key pose has it, not scaled to the target bodies' surfaces; the trunk's joint angles are not refitted to the adapted spine, only shifted with the hips.

    adapt