Couple learning for semi-supervised sound event detection (2021-10-12T00:00:00.000000Z)

TL;DR

An effective Couple Learning method that combines a well-trained model and a Mean Teacher model that increases strongly- and weakly-labeled data and reduces the noise impact in the pseudo-labels introduced by detection errors is proposed.

Abstract

The recently proposed Mean Teacher method, which exploits large-scale unlabeled data in a self-ensembling manner, has achieved state-of-the-art results in several semi-supervised learning benchmarks. Spurred by current achievements, this paper proposes an effective Couple Learning method that combines a well-trained model and a Mean Teacher model. The suggested pseudo-labels generated model (PLG) increases strongly- and weakly-labeled data to improve the Mean Teacher method-s performance. Moreover, the Mean Teacher-s consistency cost reduces the noise impact in the pseudo-labels introduced by detection errors. The experimental results on Task 4 of the DCASE2020 challenge demonstrate the superiority of the proposed method, achieving about 44.25% F1-score on the public evaluation set, significantly outperforming the baseline system-s 32.39%. At the same time, we also propose a simple and effective experiment called the Variable Order Input (VOI) experiment, which proves the significance of the Couple Learning method. Our developed Couple Learning code is available on GitHub.

Authors

Xiangdong Wang

2 papers

Ruijie Tao

5 papers

Long Yan

1 papers

Couple learning for semi-supervised sound event detection

TL;DR

Abstract

Authors

References25 items

Acoustic Event Detection with Classifier Chains

An Effective Mutual Mean Teaching Based Domain Adaptation Method for Sound Event Detection

Improving weakly supervised sound event detection with self-supervised auxiliary tasks

Audiovisual transfer learning for audio tagging and sound event detection

Sound Event Detection by Consistency Training and Pseudo-Labeling With Feature-Pyramid Convolutional Recurrent Neural Networks

Comparison of Deep Co-Training and Mean-Teacher Approaches for Semi-Supervised Audio Tagging

Forward-Backward Convolutional Recurrent Neural Networks and Tag-Conditioned Convolutional Neural Networks for Weakly Labeled Semi-supervised Sound Event Detection

Sound Event Detection and Separation: A Benchmark on Desed Synthetic Soundscapes

Gated Multi-Head Attention Pooling for Weakly Labelled Audio Tagging

DCASENET: An Integrated Pretrained Deep Neural Network for Detecting and Classifying Acoustic Scenes and Events

Intra-Utterance Similarity Preserving Knowledge Distillation for Audio Tagging

Training Sound Event Detection on a Heterogeneous Dataset

Pseudo Labeling and Negative Feedback Learning for Large-Scale Multi-Label Domain Classification

Multi-Level Fusion based Class-aware Attention Model for Weakly Labeled Audio Tagging

Guided Learning Convolution System for DCASE 2019 Task 4

Sound Event Detection in Domestic Environments with Weakly Labeled Data and Soundscape Synthesis

Sound Event Detection in the DCASE 2017 Challenge

Learning to Recognize Transient Sound Events using Attentional Supervision

Weight-averaged consistency targets improve semi-supervised deep learning results

ZHENG USTC TEAM’S SUBMISSION FOR DCASE2021 TASK4 - SEMI-SUPERVISED SOUND EVENT DETECTION Technical

MEAN TEACHER WITH SOUND SOURCE SEPARATION AND DATA AUGMENTATION FOR DCASE 2020 TASK 4

Joint Training of Guided Learning and Mean Teacher Models for Sound Event Detection

Dcase challenge2020 task-4

TRAINING METHOD USING CLASS-FRAME PSEUDO LABEL FOR WEAKLY LABELED DATASET ON DCASE2019

MEAN TEACHER CONVOLUTION SYSTEM FOR DCASE 2018 TASK 4

Field of Study

Venue Information

Name

Type

URL

Alternate Names