High-recall-rate weak-annotation sound event detection method

Abstract

The invention discloses a high-recall-rate weak-annotation sound event detection method, and the method comprises the steps: setting a neural network and training data corresponding to deep learning;initializing a loss function as cross entropy loss, and adding a plurality of groups of dice losses with different weights, wherein the higher the positive sample proportion is, the larger the required weight is; training, testing and observing experimental results of only using cross entropy loss and increasing a plurality of groups of dice loss with different weights; adjusting a weight hyper-parameter in the loss, and re-performing a plurality of groups of dice loss weight values; carrying out the loop iteration to find out the best effect to complete training, and obtaining a final loss function; applying the final loss function to a neural network detection model, applying the obtained model to a sound event detection system, and obtaining packet-level prediction and frame-level prediction of a sound event through a neural network classifier. According to the method, the problem of non-uniform sample distribution caused by one-to-many classification generally adopted in sound event detection can be solved, and the F2 score paying more attention to the recall rate is effectively improved.

Description

technical field [0001] The invention belongs to the technical field of sound event detection, and in particular relates to a sound event detection method with high recall rate and weak annotation. Background technique [0002] The purpose of sound event detection (Sound event detection, SED) is to identify the sound events that occur in an audio clip, and detect the start and end times of the events. Since the 20th century, with the development of digital signal processing technology, it has become possible to use machines to realize operations such as speech recognition and music processing. With the passage of time, speech recognition technology has become more and more mature, and people have studied more auditory information more extensively. More and more applications, such as environmental sound perception and multimedia information retrieval, have put forward higher requirements for sound event detection technology. demand. Different from tasks such as audio classif...

AI Technical Summary

Problems solved by technology

[0009] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide a high-recall rate weakly labeled sound event detection method, aiming at the situation of unbalanced samples, so that the final performance of the model is better on the F2 score that pays more attention to the recall rate performance, get more accurate sound event detection results

Images