Improving Underwater Acoustic Classification Through Learnable Gabor Filter Convolution and Attention Mechanisms

2025-12-08

AI Summary

Remote detection and classification of underwater acoustic targets face challenges including the complexity of ship-radiated noise, scarcity of labeled data, and poor generalization across diverse acoustic environments. To address these, this paper proposes GSE ResNeXt—a novel deep architecture featuring a learnable 2D adaptive bandpass Gabor convolutional layer that replaces conventional fixed-bandwidth filters; integrated Squeeze-and-Excitation channel-wise attention to enhance discriminative feature learning and training stability; and the first systematic analysis revealing the critical impact of sensor-target distance on temporal generalization performance. Evaluated on three progressively complex underwater acoustic classification tasks, GSE ResNeXt achieves significantly higher accuracy than Xception, ResNet, and MobileNetV2, reduces training time by 28%, and demonstrates markedly improved robustness in cross-SNR and cross-waterbody generalization.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Remotely detecting and classifying underwater acoustic targets is critical for environmental monitoring and defence. However, the complex nature of ship-radiated and environmental underwater noise poses significant challenges to accurate signal processing. While recent advancements in machine learning have improved classification accuracy, issues such as limited dataset availability and a lack of standardised experimentation hinder generalisation and robustness. This paper introduces GSE ResNeXt, a deep learning architecture integrating learnable Gabor convolutional layers with a ResNeXt backbone enhanced by squeeze-and-excitation attention mechanisms. The Gabor filters serve as two-dimensional adaptive band-pass filters, extending the feature channel representation. Its combination with channel attention improves training stability and convergence while enhancing the model's ability to extract discriminative features. The model is evaluated on three classification tasks of increasing complexity. In particular, the impact of temporal differences between the training and testing data is explored, revealing that the distance between the vessel and sensor significantly affects performance. Results show that, GSE ResNeXt consistently outperforms baseline models like Xception, ResNet, and MobileNetV2, in terms of classification performance. Regarding stability and convergence, the addition of Gabor convolutions in the initial layers of the model represents a 28% reduction in training time. These results emphasise the importance of signal processing strategies in improving the reliability and generalisation of models under different environmental conditions, especially in data-limited underwater acoustic classification scenarios. Future developments should focus on mitigating the impact of environmental factors on input signals.

Problem

• Classifying underwater acoustic targets in noisy environments
• Addressing limited dataset availability in acoustic classification
• Improving model generalization across varying environmental conditions

Innovation

• Learnable Gabor filters for adaptive band-pass filtering
• Squeeze-and-excitation attention for channel feature enhancement
• GSE ResNeXt architecture combining both for faster convergence