Neural Edge Histogram Descriptors for Underwater Acoustic Target Recognition

2025-03-17

AI Summary

To address the reliance on computationally intensive pre-trained models and poor cross-domain generalization in underwater acoustic target recognition, this paper proposes the lightweight Neural Edge Histogram Descriptor (NEHD)—the first adaptation of an image-domain NEHD method to passive sonar signal classification. NEHD jointly leverages time-frequency representation, statistical texture analysis, and structural texture modeling to extract robust, discriminative features without requiring pre-training. Under stringent computational constraints typical of edge devices, NEHD achieves recognition accuracy comparable to large pre-trained models across multiple underwater acoustic datasets, while reducing inference overhead by over an order of magnitude. Moreover, it significantly enhances cross-domain generalization and deployment robustness. This work establishes a highly efficient and practical paradigm for edge-based underwater target recognition.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Numerous maritime applications rely on the ability to recognize acoustic targets using passive sonar. While there is a growing reliance on pre-trained models for classification tasks, these models often require extensive computational resources and may not perform optimally when transferred to new domains due to dataset variations. To address these challenges, this work adapts the neural edge histogram descriptors (NEHD) method originally developed for image classification, to classify passive sonar signals. We conduct a comprehensive evaluation of statistical and structural texture features, demonstrating that their combination achieves competitive performance with large pre-trained models. The proposed NEHD-based approach offers a lightweight and efficient solution for underwater target recognition, significantly reducing computational costs while maintaining accuracy.

Problem

• Recognize underwater acoustic targets using passive sonar.
• Reduce computational costs in target recognition tasks.
• Improve domain adaptation for acoustic signal classification.

Innovation

• Adapts NEHD for passive sonar signal classification
• Combines statistical and structural texture features
• Lightweight, efficient underwater target recognition solution