Springs

Mask Detection Tool

About Project
Our team created a machine learning-based web application that allows tracking incoming and leaving customers to ensure they follow pandemic restrictions (wearing the covid-mask on a face).

The tool is developed especially for European retail stores and warehouses that suffered a lot during the Covid-19 pandemic.
Client's goal
The main client's goal was to simplify compliance with indoor safety precautions for store owners to lower health risks and corresponding operational expenses.

At the same time, the client was searching for a team with great ML, Pytorch, and Python expertise.
About Project
Our team created a machine learning-based web application that allows tracking incoming and leaving customers to ensure they follow pandemic restrictions (wearing the covid-mask on a face).

The tool is developed especially for European retail stores and warehouses that suffered a lot during the Covid-19 pandemic.
Client's goal
The main client's goal was to simplify compliance with indoor safety precautions for store owners to lower health risks and corresponding operational expenses.

At the same time, the client was searching for a team with great ML, Pytorch, and Python expertise.
Product
Springs engineers developed a Python-based web application that allows users to connect to in-store monitoring cameras via IP to analyze the captured video in real-time.

The analysis is based on the ML identification of people who wear or don't wear safety masks to diminish the risk of catching Covid-19 within a facility.
Result
As a type of in-store analytics software the app addresses new challenges that the retail and wholesale scopes faced due to the pandemic.

The solution provides reports on the current number of customers and detailed data concerning each entry and exit of each of them from the embraced room. The full version will allow selecting object classification types to set custom tracking modes.
Just take a look

Desktop

Technologies
1/ 1
|
01Java
02Vue
03Python
04Tensorflow