A method for simultaneous processing of sound source identification and distance estimation

By using a multi-task learning model based on transfer learning, combined with fast Fourier transform and normalization, the problem of underwater sound source identification and distance estimation cannot be achieved simultaneously is solved. This enables accurate identification of underwater sound source categories and accurate distance estimation, improving the model's generalization ability and training efficiency.

CN118261237BActive Publication Date: 2026-06-02NORTHWESTERN POLYTECHNICAL UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
NORTHWESTERN POLYTECHNICAL UNIV
Filing Date
2024-04-07
Publication Date
2026-06-02

AI Technical Summary

Technical Problem

In existing technologies, underwater sound source identification and distance estimation cannot be achieved simultaneously, which limits the accuracy and usability of identification and localization tasks, and also results in a long training time.

Method used

By employing a multi-task learning model based on transfer learning, combined with fast Fourier transform and normalization processing, a method capable of simultaneously identifying underwater sound source categories and estimating sound source distances is constructed through the integration of transfer learning and multi-task learning models.

Benefits of technology

It achieves accurate identification of underwater sound source types and accurate distance estimation, improves the model's generalization ability and training efficiency, and is suitable for scenarios with only a small amount of underwater acoustic data.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118261237B_ABST
    Figure CN118261237B_ABST
Patent Text Reader

Abstract

The present disclosure provides a method for synchronous processing of sound source identification and distance estimation, comprising: obtaining underwater acoustic data of a target sea area; performing data preprocessing on the underwater acoustic data of the target sea area to obtain target feature data; obtaining a multi-task learning model based on transfer learning trained in advance; inputting the target feature data into the multi-task learning model based on transfer learning, and determining the sound source category and sound source distance value corresponding to the target feature data through the output result of the multi-task learning model based on transfer learning. Thus, by combining the multi-task learning model and the transfer learning model, a multi-task learning model based on transfer learning is constructed, which can simultaneously identify the underwater sound source category and the underwater sound source distance estimation.
Need to check novelty before this filing date? Find Prior Art