Discriminative Singular Spectrum Analysis for Bioacoustic Classification
Discriminative Singular Spectrum Analysis for Bioacoustic Classification
复制标题
DOI：
10.21437/interspeech.2020-2134
复制
发表时间：
2020-10
期刊：
影响因子：
--
通讯作者：
B. Gatto;E. M. Santos;J. Colonna;Naoya Sogi;L. S. Souza;K. Fukui
中科院分区：
文献类型：
--
作者：
B. Gatto;E. M. Santos;J. Colonna;Naoya Sogi;L. S. Souza;K. Fukui
生物声信号的分类是生态监测的基础工作。然而，这项任务包括几个挑战，如不均匀的信号长度，环境噪声和稀缺的训练数据。为了解决这些挑战，我们提出了一种判别机制来分类生物声信号，它不需要大量的训练数据，并处理不均匀的信号长度。所提出的方法依赖于将输入信号变换到由奇异谱分析（SSA）产生的子空间。然后，子空间之间的差异被用作判别空间，提供判别特征。该公式允许一个无分割的方法来表示和分类生物声学信号，以及从SSA继承的高度紧凑的描述符。我们使用包含各种生物声学信号的具有挑战性的数据集来验证所提出的方法。
Classifying bioacoustic signals is a fundamental task for ecological monitoring. However, this task includes several challenges, such as nonuniform signal length, environmental noise, and scarce training data. To tackle these challenges, we present a discriminative mechanism to classify bioacoustic signals, which does not require a large amount of training data and handles nonuniform signal length. The proposed method relies on transforming the input signals into subspaces generated by the singular spectrum analysis (SSA). Then, the difference between the subspaces is used as a discriminative space, providing discriminative features. This formulation allows a segmentation-free approach to represent and classify bioacous-tic signals, as well as a highly compact descriptor inherited from the SSA. We validate the proposed method using challenging datasets containing a variety of bioacoustic signals.