SDLGASJul 6, 2021

Self-training with noisy student model and semi-supervised loss function for dcase 2021 challenge task 4

arXiv:2107.02569v114 citations
Originality Synthesis-oriented
AI Analysis

This work addresses sound event detection for audio processing applications, but it appears incremental as it builds on existing mean-teacher and noisy student techniques.

The paper tackles polyphonic sound event detection for the DCASE 2021 Challenge Task 4 by proposing a two-stage method combining a mean-teacher model and a self-training noisy student model with semi-supervised loss, achieving evaluation on the validation set and using ensemble models for final selection.

This report proposes a polyphonic sound event detection (SED) method for the DCASE 2021 Challenge Task 4. The proposed SED model consists of two stages: a mean-teacher model for providing target labels regarding weakly labeled or unlabeled data and a self-training-based noisy student model for predicting strong labels for sound events. The mean-teacher model, which is based on the residual convolutional recurrent neural network (RCRNN) for the teacher and student model, is first trained using all the training data from a weakly labeled dataset, an unlabeled dataset, and a strongly labeled synthetic dataset. Then, the trained mean-teacher model predicts the strong label to each of the weakly labeled and unlabeled datasets, which is brought to the noisy student model in the second stage of the proposed SED model. Here, the structure of the noisy student model is identical to the RCRNN-based student model of the mean-teacher model in the first stage. Then, it is self-trained by adding feature noises, such as time-frequency shift, mixup, SpecAugment, and dropout-based model noise. In addition, a semi-supervised loss function is applied to train the noisy student model, which acts as label noise injection. The performance of the proposed SED model is evaluated on the validation set of the DCASE 2021 Challenge Task 4, and then, several ensemble models that combine five-fold validation models with different hyperparameters of the semi-supervised loss function are finally selected as our final models.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes