Artificial intelligence acoustic inspection system
AI Square
State-of-the-art acoustic inspection algorithm developed in-house with domestic technology
AI² is an advanced acoustic inspection algorithm developed independently with domestic technology and is an acoustic detection, tracking and analysis system and solution using a multi-channel microphone sensor.
Intelligent acoustic inspection system Smart, industrial equipment failure diagnosis and monitoring system Tfoi&Tfos, urban unmanned aerial vehicle (UAV) detection system AirScope
It is divided into three categories.
Acoustic Inspector ×Artificial Iintelligence
AI² is a state-of-the-art acoustic inspection developed independently with domestic technology.
Algorithm using multi-channel microphone sensor
Acoustic detection, tracking and analysis systems and solutions.
Intelligent acoustic tester, industrial equipment failure diagnosis and monitoring system,
Urban Unmanned Aerial Vehicle (UAV) Detection System
It is divided into three categories.
AcousticInspector
× ArtificialIintelligence
AI²’s learning data construction process
Check out AI²’s five-step learning data construction process.
STEP 1
Source data collection
from the measurement object through a multi-channel microphone sensor.
Abnormal noise generated is collected by band, and
From the camera within the multi-channel microphone sensor
With an image of the measured objectSecured with original data
STEP 2
data design
Analyze abnormal noise for each secured band and
To analyze the location and size of the acquired sound
design classification criteria
STEP 3
data labeling
To analyze the location and size of the acquired sound
Using classification criteriaLabeling by dataproceed
STEP 4
Data set construction
Using labeled data, abnormal noise coordinates and
By integrating image information for each measurement object,
To detect accurate location informationData set construction
STEP 5
A.I. modeling
Build an artificial intelligence (A.I) modelTo
Different types such as linear or logistic regression
Use it as learning data using algorithms
A.I-based acoustic data learning & definition,
Enhancing detection power through deep learning
Based on machine learning normal acoustic dataWow Abnormal acoustic dataare learned and defined respectively, and
Accumulated data can be used to enhance detection power through deep learning.
Immediately upon detection of abnormal dataReal-time visualization algorithmThrough the location of the abnormal noise source
provided to users.
Acoustic data collected through a multi-channel microphone sensor isDecision TreeIt is classified through abnormal noise, environmental noise, and irregular noise based on normal sound data.
A complex node is created, and the resulting class is collected and judged.
machine visionIt is a cutting-edge detection algorithm that determines the operating state of the product and the physical characteristics of the inspection object, and allows complex judgment by combining it with acoustic data.
capable of making complex judgments
Advanced detection algorithm
Complex
able to judge
Advanced detection algorithm
Via AI²in 3D spaceLocation of abnormal noise occurring based on videostandard colorNot only can it be implemented through a discrimination algorithm, it can be used as a targeting technique by removing environmental noise and combining possible acoustic data for each space with a learning algorithm. Auto Gainsecond
You can use it to enhance detection power.
using auto gain
Advancement of detection power
