Sound event detection using weakly-labeled semi-supervised data with GCRNNS, VAT and Self-Adaptive Label Refinement

TL;DR

This paper introduces a novel semi-supervised sound event detection method using GCRNNs, VAT, and self-adaptive label refinement, achieving significant improvements in detection accuracy on weakly labeled data.

Contribution

The paper proposes a new approach combining gated convolutional recurrent neural networks, virtual adversarial training, and self-adaptive label refinement for weakly labeled sound event detection.

Findings

Achieved a macro F-score of 34.6% on the DCASE 2018 challenge.

Improved detection performance by 20.5% over baseline.

Effectively utilized unlabeled data for training.

Abstract

In this paper, we present a gated convolutional recurrent neural network based approach to solve task 4, large-scale weakly labelled semi-supervised sound event detection in domestic environments, of the DCASE 2018 challenge. Gated linear units and a temporal attention layer are used to predict the onset and offset of sound events in 10s long audio clips. Whereby for training only weakly-labelled data is used. Virtual adversarial training is used for regularization, utilizing both labelled and unlabeled data. Furthermore, we introduce self-adaptive label refinement, a method which allows unsupervised adaption of our trained system to refine the accuracy of frame-level class predictions. The proposed system reaches an overall macro averaged event-based F-score of 34.6%, resulting in a relative improvement of 20.5% over the baseline system.