Open Collections

Liu, Xiangrui

University of British Columbia

Marine mammals and their ecosystem face significant threats from, for example, military active sonar and marine transportation. To mitigate this harm, early detection and classification of marine mammals are essential. Recent solutions involve spectrogram comparison and machine learning. However, the solutions show weaknesses in efficiency. Therefore, we propose a novel knowledge distillation framework, named XCFSMN, for this problem. We construct a teacher model that fuses the features extracted from an X-vector extractor, a DenseNet, and a Cross-Covariance attended compact Feed-Forward Sequential Memory Network (cFSMN). The teacher model transfers knowledge to a simpler cFSMN model through a temperature-cooling strategy for efficient learning. Compared to multiple convolutional neural network backbones and transformers, the proposed framework achieves state-of-the-art efficiency and performance. The improved model size is approximately 25 times smaller and the inference time is 27 times shorter on average without affecting the model’s accuracy significantly.

Text

2024-01-11

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/87221

Electrical Engineering

University of British Columbia

UBCO

http://creativecommons.org/licenses/by-nc-nd/4.0/