A streamlined multimodal emotion app that combines visual, textual, and audio inputs to provide personalized wellbeing recommendations. The current build runs in safe fallback mode for stability.
- Visual emotion detection using camera input (OpenCV-based, lightweight)
- Text emotion analysis with contextual keyword parsing
- Optional audio capture and basic emotion estimation
- Optional voice-to-text input
- Personalized wellbeing recommendations
- Photo capture with emotion overlay
- Real ML models are disabled for stability (
TensorFlow,DeepFace, etc.). - Fallback methods are used for all emotion analyses.
- The app works without any external datasets.
- Python 3.8+ on Windows
opencv-python,numpy,matplotlib,Pillow(installed viarequirements.txt)- Optional:
pyaudio,SpeechRecognition,pyttsx3for audio and voice features
Install dependencies:
pip install -r requirements.txtNotes:
- Heavy ML packages listed in
requirements.txtare optional and not used by the current build. - If
PyAudioinstallation fails on Windows:pip install pipwin pipwin install pyaudio
Start the app:
python main.pyTips:
- Click "Start Camera" to begin facial detection; use "Capture" to save a photo.
- Enter text in the input area or use "[MIC] Start Voice Input" if available.
- Use "[AUDIO] Capture Voice Emotion" to estimate emotion from recent audio.
WBRM.py/
├── main.py # Main application (unchanged logic)
├── requirements.txt # Dependencies
└── README.md # Project overview and usage
The captured_photos/ folder is created automatically when you capture photos.
- Ensure dependencies are installed:
pip install -r requirements.txt. - If camera/audio/voice features are unavailable, the app degrades gracefully.
- On Windows consoles with encoding quirks, the app uses ASCII-safe messages.
- Re-enable real models behind a configuration flag when stability allows.
- Replace fallback audio analysis with feature-based classification.
- Persist emotion history and add export options.
Provided as-is for educational and research purposes.