Technology

Sound AI That Listens and Thinks Like a Human

DEEPLY's Unique Sound AI Technology That Understands the Sound Best

The Technology Behind Listen AI

The Listen AI solution offers a comprehensive suite, including sound collection microphones, advanced sound analysis modules, servers, user-friendly dashboards, and real-time alert systems.

image
98% Accuracy Certification
We have demonstrated world-class technological excellence, backed by publications in top-tier journals and TTA certification for sound event detection, achieving a 98% accuracy (F1-score).
Real Sound-Based Precision Analysis
Our advanced analysis capabilities are powered by a dataset comprising 42 distinct sound types and unique audio data collected from real-world environments.
Accurate Sound Detection in Diverse Environments
Our technology captures target sounds with precision, even in noisy environments, and seamlessly integrates with various devices through API and SDK support.

Introducing the Listen AI Solution

We present the Listen AI solution service utilizing sound analysis AI technology.

  • Listen AI Industrial

    Machine Quality Inspection

    A solution that detects machine sounds such as motor noise or connector fastening clicks at the final production stage, ensuring precise quality inspection.

    Types of Sounds
    Motor noise, connector fastening (locking), engine vibration, etc.
    Application Areas
    Component manufacturing plants, NVH testing, power plants
    Example Installation Spaces
    EOL inspection lines, in-process quality control (IPQC)
    Accuracy
    F1-score 97% or higher
    Configuration Methods
    PLC · MES integration, on-premise server setup
  • Listen AI Safety

    Emergency · Risk Detection

    A monitoring solution designed to detect emergencies such as accidents, disturbances, and crimes through sound, enhancing safety management.

    Types of Sounds
    Screams, loud voices, disturbances, crying, collisions, glass breaking, etc.
    Application Areas
    Urban crime prevention/control, public/private physical security, health
    Example Installation Spaces
    Children's zones, hospitals, parking lots, restrooms
    Accuracy
    F1-score 97%
    Configuration Methods
    Edge / On-Prem / Cloud
  • Custom Solution

    Custom Sound Detection

    A fully tailored solution that enables customers to detect and analyze specific sounds of their choice.

    Types of Sounds
    No Restrictions
    Application Areas
    No Restrictions
    Example Installation Spaces
    No Restrictions
    Accuracy
    Sound Verification Required
    Configuration Methods
    Edge / On-Prem / Cloud

Solution Operation Method

Edge Analysis Server Method

Effective local data processing saves network bandwidth and reduces latency, providing a higher quality user experience.

image
Product Type
Screams, loud voices, disturbances, crying, collisions, glass breaking, etc.
Product Components
Microphone, Edge Analysis Server, Central Management Server, Monitoring System (CMS)
Monitoring Software
Provides Listen AI Monitoring System
Requirements
Ready to use immediately without the need for a dedicated development team or integration work

Features and Benefits

  • AI computation processing within edge devices
    By integrating sound AI analysis technology into edge devices, all sound AI computations can be processed internally, allowing for effective local data processing. We recommend AI analysis functions that impose a light load suitable for the computational performance of edge devices.
  • 70% Bandwidth Savings Compared to Traditional Methods
    Each edge device consumes approximately 280 kbps of bandwidth when sending audio clips to the server. As the number of devices increases, bandwidth requirements also rise; however, the edge device system significantly reduces bandwidth requirements to around 1 kbps.
  • Privacy Protection
    Raw audio data remains on the local edge analysis server, and only the analysis results are sent to the central management server. This approach guarantees maximum privacy.

On-Premise Method

AI analysis is processed on the central analysis server, ensuring that data is not exposed externally, with advantages of low network costs and latency.

image
Product Type
Friction sounds, engine noises, abnormal noises, etc.
Product Components
Microphone, Central Analysis Server, Monitoring System (CMS)
Monitoring Software
Provides Listen AI Monitoring System
Requirements
Ready to use immediately without the need for a dedicated development team or integration work

Features and Benefits

  • Powerful AI Analysis Capabilities
    With significantly more powerful computational capabilities compared to edge analysis servers, we can apply more diverse and accurate AI analysis functions using large-scale AI models. The on-premise method is particularly suitable for building large-scale systems.
  • Data Leakage Prevention
    Unlike other solutions, the on-premise method ensures that sound data is transmitted solely within the internal network and does not traverse the internet. Data collected through microphones is sent directly to the internal central server, guaranteeing it remains secure and protected from external exposure.
  • Cost-Effective Network Operations
    By leveraging efficient data transmission with minimal network latency, this solution eliminates costs associated with transmitting sound data over the internet. This results in economical and optimized network operation.

On-Device Method

Listen AI’s deep learning solution can be integrated directly into customer devices equipped with built-in microphones, such as security cameras.

image
Product Type
Embedded C++ SDK
Product Components
Software SDK (Android / Embedded C++ / Python)
Monitoring Software
Requires Client-Side Implementation
Requirements
Detailed information file available for download