05Computer vision / ML systemsUniversity team project

Traffic Sign Detection System

A full-stack traffic-sign recognition system connecting a React interface, Node.js API, Python ML service, and database-backed dashboards.

Public repository

The context

A team-built system for traffic-sign recognition.

This university team project connects traffic-sign recognition to a web application with authentication and role-based dashboards. The complete system includes React, Node.js/Express, Python/Flask, TensorFlow/Keras, and CockroachDB. My contribution focused on the application layers around the detection workflow; the full system and ML model are shared project scope, not work I independently built or trained.

Application screenshots

Traffic Sign AI system overview describing image uploads, predictions, request status, detection history, and exports — System overview: a guide to the application's main features.

Traffic Sign AI request flow showing upload, validation, model processing, saving the prediction, and notification — Request flow: the steps from uploading an image to receiving a prediction.

My contribution

Where I contributed

01
Contributed to frontend setup and implementation.
02
Worked on authentication and dashboard functionality.
03
Helped integrate the application layers around the detection workflow.

Architecture

How the work moves

01React image upload

02Node.js / Express API

03Flask model inference

04Stored results + dashboards

Team project scope

The complete system

01React interface for image upload and detection history

02Node.js/Express API connecting the application services

03Python/Flask inference service using TensorFlow/Keras

04CockroachDB-backed, role-based dashboards

Honest evaluation

Limits and trade-offs

This is a team university project with shared contributions.
My listed contribution focuses on frontend, authentication, dashboard, and integration work; it does not claim independent ownership of the full system or model training.
No personal performance metrics are presented. The system scope is documented in the repository README, not independently benchmarked here.
Running the complete system requires database and service configuration plus a compatible trained model; the application was not run as part of this portfolio update.

Inspect the work

Stack and reproduction

1Follow the public repository README to install frontend, backend, and Python ML service dependencies.
2Configure the documented environment variables, CockroachDB database, and Redis service, then apply the database migrations.
3Provide a compatible trained Keras model through MODEL_PATH and start the Flask service, Node.js backend, and React frontend as documented.