AI based video interviews
Pipeline: Video input (OpenCV frame sampling) -> parallel AI models (FER 2013 emotion detection, gesture analysis, vocal cue classification, ambience classification) -> JSON aggregation -> React dashboard with annotated output video.
- Backend: Python, SQLite, service-oriented routes (users, cases)
- Frontend: React (Create React App), StartX institution dashboard + Interview Practice candidate flow
- AI/ML: FER 2013 emotion detection, gesture analysis, vocal cue classification, ambience/environment classification
- Dataset: FER 2013 + custom gesture/vocal/ambience datasets
See llms.txt for an AI-readable project summary and cross-links to related AI/ML work.