Blink rate detector

Inspiration

In modern digital environments, tracking personal wellness and understanding how specific sessions or activities impact focus, relaxation, and physical vitals remains a challenge. We were inspired to build Chrono to bridge the gap between digital content and non-invasive, real-time health monitoring using accessible computer vision technology.

What it Does

Chrono is a full-stack health-monitoring application featuring:

Contactless Vitals Monitoring: Uses standard device cameras to measure pulse rate, breathing rate, and facial-expression scores in real time—no wearables required.
Session & Vitals Correlation: Tracks how your body responds during guided audio or relaxation sessions, comparing early versus late session averages to highlight subtle physiological shifts.
Robust Data & Persistence: Securely stores aggregated session readings in an isolated SQLite database layer under individual user accounts.
Interactive Insights & Exports: Provides users with clean trend charts, session summaries, and exportable reports (.txt, .csv, .json) to track their wellness journey over time.

How We Built It & Deployed

Frontend Layer: Built with Streamlit Community Cloud to deliver a responsive, intuitive user interface for session selection, user authentication, and interactive data visualization.
Camera Monitoring Backend: Powered by a robust Node.js service successfully deployed and hosted live on Render, handling real-time frame streaming and secure SDK integration.
End-to-End Architecture: Seamlessly bridges multiple runtime environments—connecting the Streamlit frontend, the Render-hosted camera monitor, and the SQLite storage backend.
Advanced Signal Processing: Filters incoming biometric data for high-confidence stable samples (confidence $\ge 50\%$) to ensure clean, reliable early-versus-late comparisons.

Challenges We Faced

Bridging a real-time computer vision pipeline across a distributed cloud architecture—connecting a Streamlit frontend to a Render-hosted camera backend while managing robust session watchdogs, frame-rate handling, and signal filtering under strict deployment constraints—was our primary technical achievement.

Built With

db
gemini
mediapipe
node.js
python
render
streamlit

Submitted to

OxHack 2026

Created by

Focused on full-stack architecture, cloud deployment, and infrastructure integration. Led the deployment of the Node.js camera monitoring backend on Render (handling port binding, dynamic routing, and environment configuration) and bridged it with the Streamlit frontend. Resolved cross-service communication hurdles to ensure seamless real-time browser frame streaming and secure data persistence.

Aishwarya Murugappan
Chen Yuan
ThomasH0001
Issaka Al-Hassan