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

arXiv — cs.LGThursday, December 18, 2025 at 5:00:00 AM
  • A new study has introduced GSE ResNeXt, a deep learning architecture that enhances underwater acoustic target classification by integrating learnable Gabor convolutional layers with a ResNeXt backbone and squeeze-and-excitation attention mechanisms. This innovation addresses the challenges posed by complex underwater noise and limited datasets, improving the model's ability to extract discriminative features.
  • This development is significant as it aims to improve the accuracy and robustness of underwater acoustic classification, which is essential for environmental monitoring and defense applications. The integration of advanced machine learning techniques could lead to more reliable detection and classification of underwater targets.
  • The introduction of GSE ResNeXt reflects a broader trend in artificial intelligence where deep learning architectures are increasingly being tailored to specific challenges across various domains. Similar advancements in medical imaging and other fields highlight the ongoing efforts to enhance classification systems, demonstrating the potential for machine learning to address complex real-world problems.
— via World Pulse Now AI Editorial System

Was this article worth reading? Share it

Recommended apps based on your readingExplore all apps
Continue Readings
Arithmetic-Intensity-Aware Quantization
PositiveArtificial Intelligence
A new framework called Arithmetic-Intensity-Aware Quantization (AIQ) has been introduced to optimize the performance of neural networks by selecting per-layer bit-widths that enhance arithmetic intensity while minimizing accuracy loss. This method has shown a significant increase in throughput and efficiency on models like ResNet-20 and MobileNetV2, outperforming traditional quantization techniques.
One-Cycle Structured Pruning via Stability-Driven Subnetwork Search
PositiveArtificial Intelligence
A new one-cycle structured pruning framework has been proposed, integrating pre-training, pruning, and fine-tuning into a single training cycle, which aims to enhance efficiency while maintaining accuracy. This method identifies an optimal sub-network early in the training process, utilizing norm-based group saliency criteria and structured sparsity regularization to improve performance.

Ready to build your own newsroom?

Subscribe to unlock a personalised feed, podcasts, newsletters, and notifications tailored to the topics you actually care about