Adaptive acoustic event classification methods and devices, and computer-readable invariant means.

VN10061749BActive Publication Date: 2026-09-22QUALCOMM INC
0 Cites 0 Cited by

Patent Information

Application Number
VN1202303175
Authority / Receiving Office
VN · VN
Patent Type
Patents
Current Assignee / Owner
Priority Date
2020-11-24
Filing Date
2021-11-19
Publication Date
2026-09-22
Estimated Expiration
2041-11-19

AI Technical Summary

Technical Problem

Existing sound event classification systems are domain-specific and difficult to update for new sound classes or variants, requiring significant computational resources and labeled training data, leading to inefficiencies in adapting to changing acoustic environments.

Method used

The use of transfer learning techniques to update sound event classification models by determining if unrecognized sounds are due to drift or new classes, with model updates based on scene data and user input, allowing for adaptation with fewer resources and without retraining from scratch.

Benefits of technology

Enables efficient adaptation of sound event classification models to recognize a broader range of sounds within classes and add new classes without performance degradation, using significantly fewer computational resources and allowing for continuous learning in resource-constrained environments.

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

Abstract

The invention relates to a method and apparatus for adaptive audio event classification. The apparatus consists of one or more processors configured to provide audio data samples to an audio event classification model. One or more processors are also configured to determine, based on the output of the audio event classification model responding to the audio data samples, whether the sound type of the audio data samples is recognized by the audio event classification model. One or more processors are also configured, based on the determination that the sound type was not recognized, to determine whether the audio event classification model corresponds to the audio scene associated with the audio data samples. One or more processors are also configured to, based on the determination that the audio event classification model corresponds to the audio scene associated with the audio data samples, store model update data based on the audio data samples.
Need to check novelty before this filing date? Find Prior Art