GARUD

GARUD: Gunshot Acoustic Recognition and Ubiquitous Detection

GARUD is a cutting-edge system designed for army troops to detect the direction of gunshots using an omnidirectional microphone, FPGA, and a machine learning (ML) model.

System Overview

1. Audio Capture and Filtering: The system first captures the ambient audio using an omnidirectional microphone. The captured audio is then filtered using an FPGA to remove noise and enhance the relevant sound signals.
2. Machine Learning Processing: The filtered audio is processed by our ML model, developed using TensorFlow, which determines the direction of arrival for the detected gunshot.
3. Reporting and Analysis: A detailed report of the detected activities is generated using Generative AI, assisting the command center in maintaining accurate records.
4. API Development and Testing: We have developed an API for this project and tested it using Postman to ensure functionality and performance.

Impact

This system significantly enhances the speed of response to gunfire incidents, enabling troops to take swift action against threats and effectively eliminate enemies.