Durare - Ai Pushup Counter

An Android application that uses face detection to count push-ups in real-time. The app detects your face position through the front camera and counts push-ups based on how close your face is to the camera.

Download

Features

• Real-time face detection using ML Kit
• Automatic push-up counting based on face position
• High sensitivity detection for easy testing
• Visual feedback with face overlay and position indicators
• Live counter display showing total push-ups completed

How It Works

The app uses the following logic to detect push-ups:

1. Down Position: When your face takes more than 50% of the screen (face close to camera)
2. Up Position: When your face takes more than 25% of the screen (face at medium distance)
3. Push-up Count: A complete push-up is counted when you transition from down to up position

Usage Instructions

1. Grant Camera Permission: Allow the app to access your front camera when prompted
2. Grant Notifcation Permission: Allow the app to prompt a notification for the rest timer
3. Position Yourself: Stand in front of your device with the front camera facing you
4. Start Push-ups:
   • Move your face close to the camera (down position)
   • Push up to move your face further from the camera (up position)
   • Repeat to count more push-ups

Technical Details

Dependencies

• ML Kit Face Detection: For real-time face detection
• CameraX: For camera preview and image analysis
• AndroidX: For modern Android development

Sensitivity Settings

• Down Threshold: 50% (face takes more than 50% of screen)
• Up Threshold: 25% (face takes more than 25% of screen)
• Frame Threshold: 3 consecutive frames required to confirm position
• Min Face Size: 0.1f (detects very small faces for high sensitivity)

Note: You can customize the Down and Up Threshold in settings/personalize

Building and Running

1. Open the project in Android Studio
2. Sync Gradle files to download dependencies
3. Connect your project to firebase and add the google_services.json file
4. Enable Authentication, Firestore Database and Ai Logic in Firebase and edit the default_web_client_id in strings.xml in app/src/main/res/values folder
5. Connect an Android device or start an emulator
6. Build and run the app
7. Grant camera permissions when prompted

Troubleshooting

• No face detected: Make sure you're in a well-lit area and your face is clearly visible
• Inconsistent counting: Try adjusting your distance from the camera
• App crashes: Ensure you have granted camera permissions
• Poor detection: Clean your camera lens and ensure good lighting

License

This project is open source and available under the GPL-3.0 license.

Screenshots

Share stats to social media stories