index.html

<!doctype html>
<html>
<head>
  <meta charset="UTF-8">
  <meta name="viewport"
        content="width=device-width, user-scalable=no, initial-scale=1.0, maximum-scale=1.0, minimum-scale=1.0">
  <title>Cube Demo</title>

  <!-- three.js -->
  <script src="https://unpkg.com/three@0.126.0/build/three.js"></script>
</head>
<body>

<!-- Starting an immersive WebXR session requires user interaction.
    We start this one with a simple button. -->
<button onclick="activateXR()">Start Cube Demo</button>
<script>
async function activateXR() {
  // Add a canvas element and initialize a WebGL context that is compatible with WebXR.
  const canvas = document.createElement("canvas");
  document.body.appendChild(canvas);
  const gl = canvas.getContext("webgl", {xrCompatible: true});

  const scene = new THREE.Scene();

// The cube will have a different color on each side.
const materials = [
  new THREE.MeshBasicMaterial({color: 0xffd1dc}),
  new THREE.MeshBasicMaterial({color: 0xb0e0e6}),
  new THREE.MeshBasicMaterial({color: 0xafeeee}),
  new THREE.MeshBasicMaterial({color: 0x98fb98}),
  new THREE.MeshBasicMaterial({color: 0xffb6c1}),
  new THREE.MeshBasicMaterial({color: 0xffdab9})
];

// Create the cube and add it to the demo scene.
const cube = new THREE.Mesh(new THREE.BoxBufferGeometry(0.5, 0.5, 0.5), materials);
cube.position.set(0, 0, -1);
scene.add(cube);

// Set up the WebGLRenderer, which handles rendering to the session's base layer.
const renderer = new THREE.WebGLRenderer({
  alpha: true,
  preserveDrawingBuffer: true,
  canvas: canvas,
  context: gl
});
renderer.autoClear = false;

// The API directly updates the camera matrices.
// Disable matrix auto updates so three.js doesn't attempt
// to handle the matrices independently.
const camera = new THREE.PerspectiveCamera();
camera.matrixAutoUpdate = false;

// Initialize a WebXR session using "immersive-ar".
const session = await navigator.xr.requestSession("immersive-ar");
session.updateRenderState({
  baseLayer: new XRWebGLLayer(session, gl)
});

// A 'local' reference space has a native origin that is located
// near the viewer's position at the time the session was created.
const referenceSpace = await session.requestReferenceSpace('local');

// Create a render loop that allows us to draw on the AR view.
const onXRFrame = (time, frame) => {
  // Queue up the next draw request.
  session.requestAnimationFrame(onXRFrame);

  // Bind the graphics framebuffer to the baseLayer's framebuffer
  gl.bindFramebuffer(gl.FRAMEBUFFER, session.renderState.baseLayer.framebuffer)

  // Retrieve the pose of the device.
  // XRFrame.getViewerPose can return null while the session attempts to establish tracking.
  const pose = frame.getViewerPose(referenceSpace);
  if (pose) {
    // In mobile AR, we only have one view.
    const view = pose.views[0];

    const viewport = session.renderState.baseLayer.getViewport(view);
    renderer.setSize(viewport.width, viewport.height)

    // Use the view's transform matrix and projection matrix to configure the THREE.camera.
    camera.matrix.fromArray(view.transform.matrix)
    camera.projectionMatrix.fromArray(view.projectionMatrix);
    camera.updateMatrixWorld(true);

    // Render the scene with THREE.WebGLRenderer.
    renderer.render(scene, camera)
  }
}
session.requestAnimationFrame(onXRFrame);

}
</script>
</body>
</html>