Cross-Domain Knowledge Transfer for Underwater Acoustic Classification Using Pre-trained Models
This work addresses data scarcity in underwater acoustic target recognition by exploring transfer learning, but it is incremental as it applies existing methods to a new domain.
The study compared ImageNet pre-trained models and pre-trained audio models for underwater acoustic target recognition, finding that ImageNet models slightly outperformed audio models in classification tasks.
Transfer learning is commonly employed to leverage large, pre-trained models and perform fine-tuning for downstream tasks. The most prevalent pre-trained models are initially trained using ImageNet. However, their ability to generalize can vary across different data modalities. This study compares pre-trained Audio Neural Networks (PANNs) and ImageNet pre-trained models within the context of underwater acoustic target recognition (UATR). It was observed that the ImageNet pre-trained models slightly out-perform pre-trained audio models in passive sonar classification. We also analyzed the impact of audio sampling rates for model pre-training and fine-tuning. This study contributes to transfer learning applications of UATR, illustrating the potential of pre-trained models to address limitations caused by scarce, labeled data in the UATR domain.