Advancing underwater acoustic target recognition via adaptive data pruning and smoothness-inducing regularization
This work addresses the challenge of recognizing ship-radiated signals in underwater environments, which is important for non-line-of-sight target detection, but it appears incremental as it builds on existing methods with specific adaptations.
The paper tackled the problem of underwater acoustic target recognition with scarce and repetitive data by proposing a data pruning and regularization strategy, resulting in a framework that significantly outperforms state-of-the-art methods in low-resource scenarios.
Underwater acoustic recognition for ship-radiated signals has high practical application value due to the ability to recognize non-line-of-sight targets. However, due to the difficulty of data acquisition, the collected signals are scarce in quantity and mainly composed of mechanical periodic noise. According to the experiments, we observe that the repeatability of periodic signals leads to a double-descent phenomenon, which indicates a significant local bias toward repeated samples. To address this issue, we propose a strategy based on cross-entropy to prune excessively similar segments in training data. Furthermore, to compensate for the reduction of training data, we generate noisy samples and apply smoothness-inducing regularization based on KL divergence to mitigate overfitting. Experiments show that our proposed data pruning and regularization strategy can bring stable benefits and our framework significantly outperforms the state-of-the-art in low-resource scenarios.