Integrated Optimization of Underwater Acoustic Ship-Radiated Noise Recognition Using Relative GMM-MAP-UBM Approach

Due to the wealth of information it contains, ship radiated noise plays a crucial role in underwater acoustic signal processing for ship identification. The traditional approach to ship radiated noise detection is extremely inefficient and inaccurate since it relies excessively on the operator's prior knowledge and experience. While doing preprocessing, feature extraction, and training the model, make sure to keep the order of significance. The application of frequency component decomposition drives data preparation. Researchers examine a wide variety of feature extraction approaches for application in ship-radiated noise detection, classifying the resulting features accordingly. Following feature extraction, with unified GMM-MAP-UBM models must be trained. Both GMM and UBM, which are considered state-of-the-art approaches, are surpassed by the proposed method. An improvement of 98.54% in accuracy was achieved after implementing the technique.