MimicEase

Face-expression smartphone control for people with severe physical disabilities

MimicEase is a free, open-source Android accessibility app that lets users with ALS, cerebral palsy, spinal cord injuries, and other motor disabilities control their smartphones entirely through facial expressions. No touch, no voice commands — just your face.

All processing happens on-device. No data ever leaves your phone.

People with severe motor impairments often cannot use touchscreens or styluses. Existing solutions are expensive, require specialized hardware, or depend on cloud services that raise privacy concerns. MimicEase turns any modern Android phone's front camera into a fully capable, privacy-respecting input device.

• 52 facial expression triggers — powered by Google MediaPipe Face Landmarker (cheek puffs, eyebrow raises, mouth gestures, jaw open, tongue out, and more)
• 3 interaction modes
  • Expression Only — fixed-coordinate taps and system actions triggered by expressions
  • Cursor Click — Bluetooth mouse moves the cursor; expressions click
  • Head Mouse — head movement controls an on-screen cursor with dwell-click support
• 35+ mappable actions — back, home, scroll, swipe, app launch, volume, media control, screen brightness, and more
• Multi-profile support — create different profiles for different postures or apps
• Global toggle — enable/disable via volume key combo, a facial expression, Quick Settings tile, or broadcast (Bixby Routines / Gemini)
• EMA smoothing — exponential moving average filter reduces jitter and false triggers
• Hold duration + cooldown — prevents accidental activations
• Fully offline — zero network access; all computation runs on-device via MediaPipe

Front Camera (ImageProxy)
      ↓
MediaPipe Face Landmarker  [52 BlendShapes + head pose matrix]
      ↓
Expression Analyzer  [EMA filter, consecutive-frame gate]
      ↓
Global Toggle Check
      ↓
Trigger Matcher  [threshold + holdDuration + cooldown]
      ↓
Action Executor  [GestureDescription, Intent, AudioManager]
      ↓
Android System  [tap, swipe, back, home, app launch, …]

In Head Mouse mode, the head-pose transformation matrix is additionally routed through HeadTracker → DwellClickController → an overlay cursor rendered on screen.

(Coming soon — contributions welcome)

Modern Android blocks direct APK installation in most cases. Use the Windows Installer below — no developer tools needed.

Bundles portable ADB and walks you through every step. Works over USB or Wi-Fi. Supports 9 languages.

1. Go to the Releases page and download MimicEase-vX.X.X-installer.zip
2. Extract the ZIP to any folder
3. Double-click install.bat ← start here
4. Select your language and follow the on-screen instructions

📄 For detailed instructions and troubleshooting, open INSTALL_GUIDE.md inside the ZIP.

⚠️ Do not double-click install.ps1 — it opens in Notepad. Always use install.bat.

Package contents after extraction:

MimicEase-vX.X.X/
├── INSTALL_GUIDE.md    ← read this if anything goes wrong
├── install.bat         ← double-click to start
├── install.ps1         ← runs automatically (do not open directly)
├── MimicEase-vX.X.X.apk
└── adb/                ← portable ADB tools (no install needed)

1. Download MimicEase-vX.X.X.apk from the Releases page
2. Enable Install from unknown sources in Android Settings → Security
3. Open the downloaded APK and install
4. Follow MimicEase onboarding and enable the accessibility service

git clone https://github.com/<your-org>/MimicEase.git
cd MimicEase

# Debug APK
./gradlew assembleDebug

# Install directly to a connected device
./gradlew installDebug

Prerequisites: Android Studio Hedgehog or newer, JDK 17+.

MimicEase follows Clean Architecture with three layers:

app/
├── data/         # Room DB · DataStore · Repository implementations
├── domain/       # Pure Kotlin models & repository interfaces (no Android deps)
├── presentation/ # Jetpack Compose screens + ViewModels
├── service/      # Background services & core logic
├── di/           # Hilt dependency injection modules
└── navigation/   # Compose Navigation graph
gameFace/
└── FaceLandmarkerHelper.java  # MediaPipe face-detection engine (Java)

Tech stack: Kotlin 2.0 · Jetpack Compose · Material 3 · Hilt · Room · DataStore · CameraX · MediaPipe · Coroutines/Flow · Gson · Timber

Every trigger is fully customizable — threshold, hold duration, cooldown, and action are all configurable per profile.

Contributions of all kinds are welcome — code, translations, testing on new devices, and accessibility feedback from users with disabilities.

1. Fork the repository.
2. Create a feature branch: git checkout -b feat/my-feature
3. Follow the developer guidelines.
4. Open a pull request with a clear description.

Please keep the domain/ layer free of Android dependencies and use System.currentTimeMillis() instead of SystemClock in service-layer code (required for unit-test compatibility).

# Unit tests (no device needed)
./gradlew :app:test

# Quick Kotlin compilation check
./gradlew :app:compileDebugKotlin

Tests cover: EMA filter logic (ExpressionAnalyzerTest), trigger matching (TriggerMatcherTest), and Gson serialization (ActionSerializerTest).

Copyright 2025 MimicEase Contributors

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

• Google Project GameFace — face-control reference implementation
• MediaPipe — on-device ML face landmarking
• CameraX — Android camera abstraction

MimicEase is built with care for people who need it most. If this project has helped you or someone you know, please consider starring the repository or sharing it with accessibility communities.