Unsupervised learning based end-to-end delayless generative fixed-filter active noise control
This work addresses the resource-intensive and potentially biased labeling process for noise control systems, offering a more practical solution for active noise control applications.
The paper tackles the need for labeled noise data in training a generative fixed-filter active noise control (GFANC) system by proposing an unsupervised-GFANC approach that uses an end-to-end differentiable system with accumulated squared error as the loss, resulting in better noise reduction performance in real experiments compared to the supervised method.
Delayless noise control is achieved by our earlier generative fixed-filter active noise control (GFANC) framework through efficient coordination between the co-processor and real-time controller. However, the one-dimensional convolutional neural network (1D CNN) in the co-processor requires initial training using labelled noise datasets. Labelling noise data can be resource-intensive and may introduce some biases. In this paper, we propose an unsupervised-GFANC approach to simplify the 1D CNN training process and enhance its practicality. During training, the co-processor and real-time controller are integrated into an end-to-end differentiable ANC system. This enables us to use the accumulated squared error signal as the loss for training the 1D CNN. With this unsupervised learning paradigm, the unsupervised-GFANC method not only omits the labelling process but also exhibits better noise reduction performance compared to the supervised GFANC method in real noise experiments.