SpeechFlow - Live Presentation Coach

An iOS app that gives real-time feedback on public speaking. Built to help users improve their delivery through speech emotion detection, pacing analysis, filler word detection, and visual posture cues.

(NOTE: Need to create your own folder called "Resources" after fetching repo, then add your own video titled "test_video.mp4" in order to analyze bundled videos)

Features

• Speech Emotion Recognition
  Uses models trained on Apple's Create ML platform to classify emotions (e.g., happy, angry, sad) from your voice.

• Pacing & Filler Word Detection (coming soon)
  Detects speaking speed and overused filler words like “um” or “like”.

• Posture & Head Movement Analysis (coming soon)
  Uses Vision APIs to provide visual feedback on eye contact and posture.

APIs Used

• AVFoundation – for audio/video capture
• SoundAnalysis – for real-time audio classification
• Create ML – for emotion recognition model training
• Vision – for facial and gesture analysis (planned)
• SwiftUI – for building the iOS UI