TnyML Hack 2024 Projects Showcase
Momentum

Team Momentum

Team Members

  • Team Lead: Akshay S Rajan - CUSAT
  • Member 2 :Rashid C A - CUSAT
  • Member 3: Naveed P N - CUSAT
  • Member 4: Amal Murali P K - CUSAT

Project Description

An intelligent and compact solution leveraging Tiny ML to detect, predict, and mitigate stampedes, ensuring public safety through real-time monitoring and analysis

The Problem

Stampedes in crowded areas often result in chaos, injuries, and fatalities, primarily due to the lack of real-time monitoring and effective crowd management solutions. Traditional methods are either slow, resource-intensive, or incapable of predicting such incidents proactively.

The Solution

We solved this problem by implementing a Machine learning model on a TinyML component named Grove vision AI Module V2.

This AI model utilizes the advanced Swift-YOLO algorithm, focusing on person recognition, and can accurately detect and tag individuals in real-time video streams. It is particularly suited for the SeeedStudio Grove Vision AI (V2) device, offering high compatibility and stability

Technical Details

Technologies/Components Used

For Software:

For Hardware:

  • Grove Vision AI Module V2
  • Seeed ESP32S3

Implementation

  • Plug and connect Grove Vision AI Module V2 to sense craft (Person Detection Model)
  • Download and install Seeed-Arduino-SSCMA Library from github (GitHub (opens in a new tab))
  • Add Library to Arduino IDE and include the header file
  • Connect ESP32S3 module with Grove Vision AI Module V2
  • Write a program to extract the person count and compare with the limit.
  • Add WIFI header file to utilize the WIFI functionality of ESP32S3 ("WiFi.h")
  • Add HTTPClient header file to make HTTP requests ("HTTPClient.h")
  • Implement Telegram notification system to Alert the organisers.
  • Write python code using pyserial to read the data from Module for further data analysis.
  • The collected data is uploaded to google sheet and graphs are created.

Installation

DevscriptHardware Tools