HackStack

Project Info

SafeSight

Devpost

This project did not submit a demo video on Devpost.

Inspiration

Annually, 2.78 million workers die due to work-related accidents and diseases, with 374 million more facing non-fatal incidents. That's 7,500 daily deaths, surpassing those from road accidents, war, and HIV/AIDS combined. (Source here) When workplace injuries do happen, many of them go unreported and workers are not compensated due to retributive fears as well as extensive time lost while filling worker comp claims.

What it does

SafeSight's AI-based technology addresses workplace injury through a combination of real-time monitoring, timely incident detection, and automated generation of accurate injury reports, ethically protecting human safety and efficiency. Our product vertically integrates and automates all parts of worker compensation, which can be summarized in 3 S’s. STOP: Forget waiting for someone to notice the danger. SafeSight reviews security footage in real-time to instantly flag potential risks and threats to workplace environments. If spotted, SafeSight instantly alerts relevant personnel, ensuring prompt preventative action is taken. SIGHT: In the unfortunate circumstance that a workplace accident occurs, SafeSight is the first to know. Our computer vision solution not only detects the incident, but also ensures that workers involved get the necessary help and treatment from the right people, whether that be medical staff, security, or management, ASAP. SETTLE: Post-event, SafeSight helps you process and understand what happened. Our systems generate a detailed report based on footage, capturing essential data that aids in both addressing current incidents and preventing future ones.

How we built it

Front-end: React (JavaScript, HTML/CSS) Back-end: Flask, Google Firebase, TensorFlow We use React.js to create the front-end of SafeSight. When user’s upload video footage, it is stored in firebase. We created a backend API using flask that then processes the video footage. We trained ML models using TensorFlow.

Analysis

Compare with all teams

View

Metric

Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.

Technology

Found in codeClaimed only
  • CSSIn code
  • FirebaseIn code
  • FlaskIn code
  • HTMLIn code
  • JavaScriptIn code
  • Next.jsIn code
  • OpenAIIn code
  • ReactIn code
  • PythonClaimed

8 of 9 appear in the indexed code. 1 claimed on Devpost could not be matched to code, which may simply mean the tool leaves no trace in the repository.

AI coding agents

No AI coding agent signals were found in this repository.

Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.

Codebase size

Source size

47 KB

Source files

26

Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.

0 stars