Acoustic Scene Classification for Context-Aware Call Security

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Automated voice-based systems lack context-awareness, making them vulnerable to malicious attacks such as impersonation and eavesdropping, particularly when users are in environments outside their approved settings and need to share sensitive information.

Innovation Solution

A computer-implemented method using acoustic scene classification to identify user environments and compare them against a whitelist, coupled with automatic speech recognition to determine the call reason, and perform security policy actions like forwarding calls to humans or multi-factor authentication when necessary.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If automated voice-based systems are used to handle telephone interactions, then productivity and ease of operation are improved, but security and reliability deteriorate due to lack of context-awareness and vulnerability to malicious attacks

Engineering Contradiction:

Improvecall handling efficiencyVSAvoidcall security

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent introduces an acoustic scene classification system as an intermediary between the automated voice-based system and the user. This intermediary analyzes the acoustic environment to determine context-awareness, enabling the system to identify potential security risks (such as public spaces or noisy environments) and trigger appropriate security measures like transferring to human agents or requesting additional verification, thereby maintaining both automation efficiency and security reliability

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements feedback mechanisms where the acoustic scene classification results are continuously monitored and used to adjust call handling decisions. The system receives feedback about the user's environment, processes this information through security policies, and dynamically adjusts the level of automation applied, creating a closed-loop control system that balances productivity and security based on real-time conditions

Inventive Principle:

Principle #23Feedback

2Reliability

If context-awareness is added to automated systems through acoustic scene classification, then security is improved, but device complexity increases

Engineering Contradiction:

Improvecall securityVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the call handling system into distinct functional modules: acoustic scene classification module, security policy evaluation module, and call routing module. Each module performs a specific function independently, making the overall complex system manageable through modular design. The acoustic scene classification is separated from the core call handling logic, allowing security enhancements without fundamentally redesigning the entire system

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The acoustic scene classification system serves multiple functions simultaneously: it provides security assessment, enables context-aware routing decisions, and can trigger different security policies based on the detected environment. This multi-functionality reduces the need for separate specialized systems, thereby limiting the increase in overall device complexity while achieving enhanced security

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Reliability

If security policy actions are performed for unwhitelisted environments, then security is improved, but ease of operation and user convenience deteriorate

Engineering Contradiction:

Improvecall securityVSAvoiduser convenience

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The system applies security policy actions selectively rather than universally. Instead of blocking all calls from unwhitelisted environments, it performs partial security checks and applies only the necessary security measures (such as multi-factor authentication or human transfer) based on the specific risk level detected. This approach provides adequate security protection while minimizing disruption to user convenience for low-risk scenarios

Inventive Principle:

Principle #16Partial or excessive action

4Measurement precision

If acoustic scene classification is implemented to identify user environments, then measurement precision of environment context is improved, but use of energy and device complexity increase

Engineering Contradiction:

Improveenvironment identification accuracyVSAvoidprocessing energy consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary acoustic scene classification at the beginning of the call to establish the user's environment context. By conducting this analysis upfront rather than continuously throughout the call, the system achieves accurate environment identification while minimizing ongoing energy consumption. The classification results are cached and reused for the duration of the call, reducing repeated processing requirements

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS12556931B2Computer-implemented method for call security

Publication Date: 2026.02.17 ELISA OYJ

AI summary

Disclosed herein is a computer-implemented method for call security, which includes: receiving a call from a user; identifying an environment of the user during the call using acoustic scene classification; comparing the identified environment to a preconfigured environment whitelist; and in response to at least the preconfigured environment whitelist not comprising the identified environment, performing at least one security policy action.