A Streamlit dashboard for reviewing self-driving clips, assigning audit outcomes, tracking failure reasons, and monitoring recollection needs. This project simulates how a data collection supervisor or quality reviewer would evaluate clip usefulness, metadata completeness, and operator performance.
- Audit new clips in the UI
- Upload CSV support
- Filter by date, auditor, operator, route, scenario, and status
- Pass / Needs Review / Fail tracking
- Failure-reason analysis
- Operator quality scorecard
- Audit queue for failed or recollection-needed clips
- Download filtered CSV exports
- Sample data fallback for testing
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
streamlit run app.py