Voice signal extraction and recognition method and device, electronic equipment and storage medium

By monitoring and analyzing the audio signals of the target environment in real time, extracting biased audio signals and using recognition models and emotional recognition models, the problem of difficult to identify and judge events in the prior art is solved, and efficient event recognition in the noisy environment is achieved.

CN119724246AActive Publication Date: 2025-03-28SHANGHAI HUITONGSHENG INTELLIGENT TECHNOLOGY CO LTD
5 Cites 0 Cited by

Patent Information

Application Number
CN202510239987.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-03-03
Publication Date
2025-03-28
Estimated Expiration
2045-03-03

AI Technical Summary

Technical Problem

The prior art is difficult to effectively identify and judge events such as loud noises and disputes, especially in noisy environments. The ambiguity and complexity of audio signals make it difficult to identify events.

Method used

By monitoring the sound signal of the target environment in real time, using the sliding window method to obtain the audio signal, and comparing it with the preset audio, extracting the biased audio signal. Then, the recognition model is used to identify the type of audio signal and the location of the sound source, and the emotional color in the audio signal is analyzed in combination with the emotion recognition model to determine the nature and location of the event.

Benefits of technology

It can extract key features from complex and fuzzy audio signals, reduce the risk of misjudgment, accurately infer the nature and location of events, improve the efficiency of event recognition, and provide a valuable reference for security management.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention provides a voice signal extraction and recognition method and device, electronic equipment and a storage medium, and belongs to the technical field of audio processing. According to the voice signal extraction and recognition method, through layer-by-layer analysis of an audio signal, deviation detection is converted into complex emotion analysis and positioning, it is ensured that even if the audio signal is complex or fuzzy, the voice signal extraction and recognition efficiency is improved; the method can still focus and extract key features step by step, reduces the risk of misjudgment, can accurately speculate the property and occurrence position of the event through comprehensive analysis of various features of the audio signal, and improves the efficiency of event property recognition through positioning and emotional color analysis of the audio signal. Clues can be found from complex and fuzzy audio signals to judge the event type, and then reference is provided for security management.
Need to check novelty before this filing date? Find Prior Art