Real-time Presentation/Interview Coach via Computer Vision and Speech Language Models(Webcam + Speech)

Structure

• src/multimodal_coach: core package
• apps/run_multimodal_coach.py: unified webcam runner
• assets/: reference video/audio/subtitles/derived pose data
• experiments/: legacy pilots and gesture FSL project
• tests/: unit tests

Run

PYTHONPATH=src python apps/run_multimodal_coach.py

API (speech feedback)

PYTHONPATH=src uvicorn multimodal_coach.api.feedback_server:app --reload