How deekshithreddykh/vaisiri-emotion-reader works
Vaisiri Emotion Reader is a desktop webcam application that captures video, detects and crops the largest face, classifies its emotion, smooths predictions, and updates a live GUI. The README documents neutral fallback, confidence display, pause/resume, and optional gallery imagery; the excerpts show the inference, smoothing, and display-update logic. Several UI and camera details are documented but not fully sampled, so uncertain internal wiring is omitted.
Main entry points: vaisiri_emotion_reader.py Desktop app
Components 13
Application runtime
- Desktop app
vaisiri_emotion_reader.py - Model initialization
vaisiri_emotion_reader.py
Vision and inference
- Camera worker
vaisiri_emotion_reader.py - Face detection
vaisiri_emotion_reader.py - Face crop
vaisiri_emotion_reader.py - Prediction smoothing
vaisiri_emotion_reader.py - Neutral fallback
vaisiri_emotion_reader.py
User experience
- Live GUI
vaisiri_emotion_reader.py - Emotion display
vaisiri_emotion_reader.py - Pause/resume controls
vaisiri_emotion_reader.py
Other components
- User
- Webcam
- DeepFace
Connections 15
- User → Desktop app: launches
- Desktop app → Live GUI: creates
- Desktop app → Model initialization: starts
- Model initialization → DeepFace: warms model
- Desktop app → Camera worker: starts worker
- Webcam → Camera worker: supplies frames
- Camera worker → Face detection: detects faces
- Face detection → Face crop: selects largest
- Face crop → DeepFace: analyzes crop
- DeepFace → Prediction smoothing: returns scores
- Prediction smoothing → Live GUI: schedules update
- Neutral fallback → Live GUI: resets display
- Live GUI → Emotion display: renders emotion
- User → Pause/resume controls: pauses or resumes
- Pause/resume controls → Desktop app: changes detection