Addressing Feature Imbalance in Sound Source Separation
This addresses feature imbalance in sound source separation, which is an incremental improvement for audio processing applications.
The paper tackles the feature preference problem in high-dimensional regression tasks like sound source separation, where neural networks overly rely on specific features, and proposes FEABASE to mitigate this by balancing features, achieving efficient data utilization.
Neural networks often suffer from a feature preference problem, where they tend to overly rely on specific features to solve a task while disregarding other features, even if those neglected features are essential for the task. Feature preference problems have primarily been investigated in classification task. However, we observe that feature preference occurs in high-dimensional regression task, specifically, source separation. To mitigate feature preference in source separation, we propose FEAture BAlancing by Suppressing Easy feature (FEABASE). This approach enables efficient data utilization by learning hidden information about the neglected feature. We evaluate our method in a multi-channel source separation task, where feature preference between spatial feature and timbre feature appears.