Investigation of Time-Frequency Feature Combinations with Histogram Layer Time Delay Neural Networks
This work addresses performance enhancement for underwater acoustic signal processing, but it is incremental as it focuses on optimizing feature combinations within an existing framework.
The paper tackled the problem of improving model performance for underwater acoustic signals by investigating different combinations of time-frequency features in a histogram layer time delay neural network, with results showing that specific feature combinations outperform single data features.
While deep learning has reduced the prevalence of manual feature extraction, transformation of data via feature engineering remains essential for improving model performance, particularly for underwater acoustic signals. The methods by which audio signals are converted into time-frequency representations and the subsequent handling of these spectrograms can significantly impact performance. This work demonstrates the performance impact of using different combinations of time-frequency features in a histogram layer time delay neural network. An optimal set of features is identified with results indicating that specific feature combinations outperform single data features.