# Not Just Seeing, but Hearing Anomalies with Sound Event Detection Module D1 Source: https://www.youtube.com/watch?v=2crLilyo_vk Recap page: https://rapidrecap.app/video/2crLilyo_vk Generated: 2026-01-30T10:03:45.42+00:00 --- ## Quick Overview The video successfully demonstrates that the new Sound Event Detection Module D1 can reliably detect and report five types of abnormal sounds, such as glass breaking, baby crying, gunshots, and smoke alarms, in real-time with high confidence (e.g., 80% for a gunshot) directly to a user's phone via email, proving its utility for ambient security monitoring. **Key Points:** - The Sound Event Detection Module D1 identifies five critical abnormal sound types: glass breaking, baby crying, gunshot, smoke alarm, and snoring. - The module processes audio events at the edge using low power, enabling continuous monitoring. - Upon detection (e.g., gunshot at 80% confidence), the system sends a real-time alert, including location and time, directly to the user's phone via email. - The demonstration confirmed that the combined vision and hearing capabilities enhance reliability compared to vision-only systems. - A promotional code J29O58TK is offered for $3 off the Sound Event Detection Module for a limited time. - The module can be integrated into various scenarios like schools, hospitals, public spaces, and residential areas for enhanced security. - The setup involves connecting the module (an ESP32 audio sensor) to a speaker and testing its response via a smartphone application. ![Screenshot at 01:39: The demonstration shows the Sound Event Detection Module D1 connected to a small speaker, while the presenter holds up a smartphone displaying an incoming email alert confirming a 'Gunshot Detected' event with 80% confidence.](https://ss.rapidrecap.app/screens/2crLilyo_vk/00-01-39.jpg) **Context:** The video presents a demonstration of a newly released hardware component, the Sound Event Detection Module D1, developed by the presenters. This module integrates sound analysis capabilities, specifically sound event detection (SED), into edge computing devices. The context is to showcase how this module can augment existing security systems, which often rely solely on visual input (cameras), by adding auditory awareness for critical safety scenarios like emergencies or break-ins. ## Detailed Analysis The video showcases the capabilities of the new Sound Event Detection Module D1, which functions as an ESP32-based audio sensor capable of real-time sound event detection at the edge with low power consumption. The presenters explain that this module detects five specific abnormal sounds: glass breaking, baby crying, gunshot, smoke alarm, and snoring. A key benefit highlighted is that combining this auditory input with existing vision AI significantly improves overall system reliability compared to vision-only solutions, especially in areas like elderly homes, schools, hospitals, and public spaces. The demonstration proves the concept by generating a gunshot sound, which the module successfully detects (e.g., 80% confidence at 02:37), and immediately sends a detailed email alert to the presenter's phone, including event type, confidence score, time, and location (02:57-03:48). The presenters conclude by noting the module is currently available for purchase on their bazaar at an attractive price, and they offer a discount code (J29O58TK) for $3 off. ### Product Introduction and Concept - The Sound Event Detection Module D1 uses sound analysis to detect anomalies like glass breaking, baby crying, and gunshots - It runs continuously at low power on the edge - It complements vision systems for enhanced security. ### Demonstration of Gunshot Detection - A gunshot sound is played, triggering an immediate alert on the presenter's phone (02:47) - The alert email provides detailed information, including event type ('Gunshot'), confidence (80%), time, and location (03:23-03:48). ### Supported Sound Events - The module is trained to recognize five distinct abnormal sound types: glass breaking, baby crying, gunshot, smoke alarm, and snoring (01:57-02:07). ### Integration and Application - The technology is suitable for various environments such as schools, hospitals, public areas, and homes (04:57-05:04) - It facilitates sound-to-IoT actions, like sending alerts directly to a phone (05:17-05:20). ### Availability and Promotion - The D1 module is available for purchase now on their bazaar (05:07) - A limited-time promotion offers $3 off using code J29O58TK (00:00 banner text). ![Screenshot at 00:00: The two presenters are seated at a workbench displaying various electronic components, including a robotic arm, a small robot figure, and hardware related to audio processing.](https://ss.rapidrecap.app/screens/2crLilyo_vk/00-00-00.jpg) ![Screenshot at 01:29: A close-up overhead view of the black circular Sound Event Detection Module PCB, connected via USB, with several blue indicator LEDs lit and one green status light active on the accompanying small development board.](https://ss.rapidrecap.app/screens/2crLilyo_vk/00-01-29.jpg) ![Screenshot at 02:47: A demonstration of the system's output where a loud sound \(implied gunshot\) is played, and the presenter holds up a smartphone showing an incoming email notification.](https://ss.rapidrecap.app/screens/2crLilyo_vk/00-02-47.jpg) ![Screenshot at 03:22: The smartphone screen is clearly visible, displaying an email alert titled 'ALERT Gunshot Detected' with a confidence level of 80% and location data.](https://ss.rapidrecap.app/screens/2crLilyo_vk/00-03-22.jpg) ![Screenshot at 04:40: The presenters confirm the successful real-time alert, with the female presenter explaining that the module can run continuously at low power at the edge.](https://ss.rapidrecap.app/screens/2crLilyo_vk/00-04-40.jpg)