SnapRise
A Flutter alarm app that forces you to prove you're awake before the alarm will stop — either by photographing a real object, or by doing hip-bend sit-ups tracked with on-device pose detection.
How it works
- Set an alarm — choose time, label, repeat days, vibration, difficulty, and dismissal challenge
- Alarm rings — the screen wakes your device with a loud alarm
- Challenge is revealed — depending on the alarm setting:
- Object detection — a spinning roulette lands on a household item; photograph it with the in-app camera
- Sit-ups — do a random target of 2–4 standing hip bends in front of the front camera
- On-device ML verifies you — Google ML Kit image labeling or pose detection (no internet needed)
- Alarm stops ✅ — only after a successful detection / completed reps
Tech Stack
| Feature | Package |
|---|
| Alarm scheduling | alarm ^5.4.1 |
| Notifications | flutter_local_notifications ^21.0.0 |
| Camera | camera ^0.12.0 |
| Object labeling | google_mlkit_image_labeling ^0.14.2 |
| Pose / sit-up tracking | google_mlkit_pose_detection ^0.14.1 |
| Permissions | permission_handler ^12.0.3 |
| Storage | shared_preferences ^2.5.5 |
App version: 2.0.0 · Display name: SnapRise
Project Structure
Flutter project lives under app/wake_up_alarm/.
app/wake_up_alarm/lib/
├── main.dart # App entry, alarm ring listener
├── models/
│ └── alarm_model.dart # Alarm data (challenge flags, difficulty)
├── services/
│ ├── alarm_scheduler_service.dart # Schedule/cancel/sync alarms
│ ├── alarm_storage_service.dart # Persist alarms to SharedPrefs
│ ├── item_selector_service.dart # Random item picker
│ └── object_detection_service.dart # ML Kit image labeling wrapper
├── screens/
│ ├── home_screen.dart # Alarm list
│ ├── add_alarm_screen.dart # Create/edit alarm + challenge mode
│ ├── ringing_screen.dart # Alarm ringing + challenge reveal
│ ├── camera_screen.dart # Camera + object detection
│ ├── situp_screen.dart # Front camera + pose-based sit-ups
│ └── result_screen.dart # Success / fail result
├── widgets/
│ └── glowing_button.dart # GlowingButton + AlarmCard
└── utils/
└── app_theme.dart # Theme + AppConstants (item list)
Setup
1. Install dependencies
cd app/wake_up_alarm
flutter pub get
2. Alarm sounds
Place WAV files in assets/sounds/ (already wired in pubspec.yaml):
assets/
sounds/
alert-alarm.wav
classic_alarm.wav
facility-alarm.wav
sound-alert-in-hall.wav
space-shooter.wav
images/
flutter:
assets:
- assets/sounds/
- assets/images/
3. Android setup
Minimum SDK is 24 (Android 7).
Needed capabilities (camera, exact alarms, wake lock, foreground service for alarm audio) are expected via the app / plugin manifests. Grant camera and alarm permissions when prompted on device.
4. iOS setup
Ensure Info.plist includes:
NSCameraUsageDescription — required for object detection and sit-up trackingUIBackgroundModes: audio — for alarm sound when backgrounded
On iOS, background alarms have limitations when the app is fully killed. The alarm package handles this as best as possible using AVAudioSession.
5. Run
cd app/wake_up_alarm
flutter run
Dismissal challenges
When creating or editing an alarm, pick one challenge under Dismissal challenge:
| Mode | What you do | How it verifies |
|---|
| Object detection | Find the roulette item and take a photo | ML Kit image labeling |
| Sit-ups | Complete 2–4 standing hip bends | ML Kit pose detection (hip/knee landmarks) |
Difficulty
Difficulty controls minAttemptsBeforeSkip (default 3, range 1–10): how many failed object-detection attempts before a skip option can appear. Configured in AppConstants and per alarm.
Detectable Items
Items use labels from ML Kit's official label map plus practical alternates.
| Item | Primary ML label | Also accepts |
|---|
| Laptop | laptop | computer, macbook, notebook, personal computer |
| Phone | mobile phone | smartphone, telephone, iphone, gadget |
| Coffee Mug | cup | coffee, mug, cappuccino, coffee cup, drinkware |
| Chair | chair | seat, furniture, stool, office chair |
| TV | television | tv, screen, monitor, display, flat screen |
| Houseplant | plant | flower, flowerpot, houseplant, tree, leaf, vegetation |
| Glasses | glasses | sunglasses, eyewear, spectacles, goggles |
| Sneakers | sneakers | shoe, footwear, boot, athletic shoe |
| Bag | bag | handbag, backpack, tote, luggage, baggage |
| Pillow | pillow | cushion, throw pillow, bedding |
To add or change items, edit AppConstants.alarmItems in lib/utils/app_theme.dart.
Detection Confidence
Default object-detection threshold: 40% (AppConstants.detectionConfidenceThreshold)
Adjust in app_theme.dart if detections are too strict or too loose.
Sit-up counting uses adaptive hip/knee position thresholds on the front camera (standing bend → down → stand).
Web Support
Flutter Web has limited support for background alarms (browser tabs must stay open). Camera works on mobile browsers. For full functionality, use the native Android/iOS build.
License
MIT
Ashish Kumar
