Unreal Movement sample overview

Updated: May 11, 2026

Overview

This sample demonstrates the OculusXRMovement plugin ecosystem, including body, eye, and face tracking integration for Meta Quest headsets. It provides three custom Animation Graph nodes, a complete retargeting system with four modes, and five pawn variants that showcase different tracking configurations.

View the sample source code on GitHub⁠.

What you will learn

Integrate body, eye, and face tracking using the OculusXRMovement plugin

Use three Animation Graph nodes: FAnimNode_OculusXRBodyTracking, FAnimNode_OculusXREyeTracking, and FAnimNode_OculusXRFaceTracking

Configure four retargeting modes (RotationAndPositions, RotationOnlyUniformScale, RotationOnly, and None)

Apply face tracking corrective shapes including combination and in-between shapes

Use per-expression modifiers to fine-tune tracking output

Generate bone mappings with the GenerateBoneMapping() utility

Visualize tracking data with five debug draw modes

Requirements

Meta Quest 2, Quest 3, or Quest 3S

Unreal Engine 5.6 with the MetaXR plugin

For build setup and configuration, see the sample README⁠.

Get started

Clone or download the sample from the GitHub repository. Open the project in Unreal Engine 5.6 with the MetaXR plugin installed. Build and deploy to your Meta Quest device. The sample includes four maps that demonstrate different tracking configurations. Start with MAP_Aura for a general overview, then explore the high-fidelity maps for advanced retargeting examples.

Explore the sample

File / Scene	What it demonstrates	Key concepts
MAP_Aura (map)	General body and face tracking overview	OculusXRMovement plugin basics, pawn setup
MAP_HighFidelity (map)	High-fidelity body tracking with retargeting	Retargeting modes, bone mapping
MAP_HighFidelity_AnimBlueprint (map)	Animation Blueprint-driven high-fidelity tracking	Animation Graph nodes, Blueprint integration
MAP_RetargetMannequinAnimBlueprint (map)	Retargeting to Mannequin skeleton	Cross-skeleton retargeting, GenerateBoneMapping()
FAnimNode_OculusXRBodyTracking	Body tracking Animation Graph node	Retargeting modes, skeletal mapping
FAnimNode_OculusXREyeTracking	Eye tracking Animation Graph node	Gaze direction, eye openness
FAnimNode_OculusXRFaceTracking	Face tracking Animation Graph node	Blend shapes, corrective shapes, per-expression modifiers

Runtime behavior

The sample provides five pawn variants, each configured with a different combination of body, eye, and face tracking. At runtime, the Animation Graph nodes receive tracking data from the OculusXRMovement plugin and apply it to the character skeleton using the configured retargeting mode.

Body tracking supports four retargeting modes that control how tracking data maps to skeletons of different proportions. Face tracking applies corrective shapes (combination and in-between) on top of base blend shapes for more realistic facial animation. Five debug draw modes visualize raw tracking data, retargeted poses, and bone mappings.

Key concepts

Two-layer architecture

The sample separates plugin code from sample code. The OculusXRMovement plugin provides low-level tracking access and the three Animation Graph nodes. The sample layer demonstrates how to configure those nodes, set up retargeting, and combine tracking types in complete pawn setups.

Retargeting modes

Four retargeting modes control how body tracking data maps to target skeletons:

RotationAndPositions: Full position and rotation retargeting for matching skeletons

RotationOnlyUniformScale: Rotation with uniform scale compensation for different skeleton sizes

RotationOnly: Rotation only, preserving target skeleton proportions

None: Raw tracking data without retargeting

Face tracking corrective shapes

Face tracking uses corrective blend shapes that activate based on combinations of base expressions:

// Combination shapes activate when multiple base shapes exceed thresholds
// In-between shapes activate at specific ranges of a base shape
// Per-expression modifiers scale individual expression values before applying

The GenerateBoneMapping() utility automates bone name matching between source tracking data and target skeleton hierarchies.