Sceneform Maintained SDK for Android

---

⚠️ This framework isn't updated anymore. Consider using Sceneview in your projects instead.

---

Sceneform Maintained is an ARCore Android SDK with Google Filament as 3D engine. This is the continuation of the Archived Sceneform

Android Augmented Reality library: AR Model Viewer, Augmented Images, Augmented Faces, Video, Depth, Cloud Anchors, Instant Placement, Light Estimation and much more in Kotlin and Java

---

Chat with us

(Please don't use it for issues, go to the GitHub Issues section instead)

---

Sceneform maintained and successor:

• Sceneform - Java Continuation : Here
• SceneView - Kotlin Successor : SceneView/sceneview-android

Differences with Google Sceneform (1.15.0, 1.16.0, 1.17.0 and 1.17.1)

• No plugin needed: Use gltf and glb 3D models files directly from assets, res/raw, local file or http/https url instead of sfa, sfb, fbx, obj,...
• Latest versions of ARCore SDK and Google Filament
• Latest gradle dependencies including AndroidX, LifecycleScope/Coroutines (SceneView only),...
• Available as mavenCentral() dependency
• Augmented Images supported
• Augmented Faces supported
• glb or gltf for 3D models (animations supported)
• hdr or ktx for Environement (IndirectLight + Skybox)
• 3D only usage supported and single dependency without ARCore (SceneView only)
• VideoNode for MediaPlayer (mp4, avi,...) Video 3D Node
• Horizontal/Vertical Plane Placement
• Depth occlusion and placement (SceneView only)
• InstantPlacement (SceneView only)
• HdrLightEstimation with adjustements between more Spectacular or Realistic
• Simple Model Viewer for basic usage, with only ArSceneView parameters (SceneView only)
• Less OpenGL knowlege needed

This repository was originaly a fork of Sceneform Copyright (c) 2021 Google Inc. All rights reserved.

Dependencies

app/build.gradle

dependencies {
     implementation "com.gorisse.thomas.sceneform:sceneform:1.23.0"
}

more...

Usage (Simple model viewer)

Update your AndroidManifest.xml

AndroidManifest.xml

<uses-permission android:name="android.permission.CAMERA" />

<application>
    ...
    <meta-data android:name="com.google.ar.core" android:value="optional" />
</application>

more...

Add the View to your layout

res/layout/main_activity.xml

<androidx.fragment.app.FragmentContainerView
    android:id="@+id/arFragment"
    android:name="com.google.ar.sceneform.ux.ArFragment"
    android:layout_width="match_parent"
    android:layout_height="match_parent" />

sample...

Edit your Activity or Fragment

src/main/java/.../MainActivity.java

override fun onCreate(savedInstanceState: Bundle?) {
    super.onCreate(savedInstanceState)

    // Load model.glb from assets folder or http url
    (supportFragmentManager.findFragmentById(R.id.arFragment) as ArFragment)
        .setOnTapPlaneGlbModel("model.glb")
}

Or

src/main/java/.../MainFragment.java

override fun onViewCreated(view: View, savedInstanceState: Bundle?) {
    super.onViewCreated(view, savedInstanceState)

    // Load model.glb from assets folder or http url
    (childFragmentManager.findFragmentById(R.id.arFragment) as ArFragment)
        .setOnTapPlaneGlbModel("https://storage.googleapis.com/ar-answers-in-search-models/static/Tiger/model.glb")
}

kotlin sample...

java sample...

Samples

glTF with animation

full video...

arFragment.setOnTapArPlaneListener { hitResult, plane, motionEvent ->
    // Create the Anchor
    arFragment.arSceneView.scene.addChild(AnchorNode(hitResult.createAnchor()).apply {
        // Create the transformable model and add it to the anchor
        addChild(TransformableNode(arFragment.transformationSystem).apply {
            renderable = model
            renderableInstance.animate(true).start()
        })
    })
}

kotlin sample project...

java sample project...

Depth Occlusion

arFragment.apply {
    setOnSessionConfigurationListener { session, config ->
        if (session.isDepthModeSupported(Config.DepthMode.AUTOMATIC)) {
            config.depthMode = Config.DepthMode.AUTOMATIC
        }
    }
    setOnViewCreatedListener { arSceneView ->
        // Available modes: DEPTH_OCCLUSION_DISABLED, DEPTH_OCCLUSION_ENABLED
        arSceneView.cameraStream.depthOcclusionMode =
            CameraStream.DepthOcclusionMode.DEPTH_OCCLUSION_ENABLED
    }
}

documentation...

sample project...

Augmented Images

sample project...

Augmented Faces

A Blender file augmented face template is available here and can be used to create your own models:

The four bones in the skeleton are moved to the corresponding ARCore poses at runtime. It is possible to attach objects to the bones or assign weight to vertices to position the objects or modify a larger mesh according to the ARCore poses. The positions of the bones in the Blender file give a general idea of how the end result will look like so the relative transform of the attached objects is important.

A face texture should be prepared as a separate file. For creating a face texture this template can be used: canonical_face.png

You can also change the model shape based on the skeleton and weights assigned to vertices, moving some small objects around the head, scaling them, etc...

You can even animate all that.

sample project...

Cloud Anchors

// Create a new anchor = the pose of which ARCore will try to resolve using the ARCore Cloud Anchor service and the provided cloudAnchorId
sceneView.session?.resolveCloudAnchor(cloudAnchorId)?.let { resolvedAnchor ->
  sceneView.scene.addChild(AnchorNode(resolvedAnchor).apply {
      addChild(VideoNode(context, MediaPlayer.create(context, R.raw.restaurant_presentation).apply {
                  this.start()
              },null)
      )
  })
}

Environment Lights

https://user-images.githubusercontent.com/83349532/144654379-a4476baf-8a22-45aa-8088-4f5e3fc7384b.mp4

// Use this mode if you want your objects to be more like if they where real
arSceneView.lightEstimationConfig = LightEstimationConfig.REALISTIC

// Use this mode if you want your objects to be more spectacular
arSceneView.lightEstimationConfig = LightEstimationConfig.SPECTACULAR

// Use this mode if you only want to apply ARCore lights colors and intensity
arSceneView.lightEstimationConfig = LightEstimationConfig.AMBIENT_INTENSITY

// Use this mode if you want to disable all ARCore light estimation lights
arSceneView.lightEstimationConfig = LightEstimationConfig.DISABLED

sample project...

Video texture

arFragment.setOnTapArPlaneListener { hitResult, plane, motionEvent ->
    // Create the Anchor
    arFragment.arSceneView.scene.addChild(AnchorNode(hitResult.createAnchor()).apply {
        addChild(VideoNode(context, MediaPlayer.create(context, R.raw.