Channel and Frequency Attention Module for Diverse Animal Sound Classification

Kyungdeuk KO  Jaihyun PARK  David K. HAN  Hanseok KO  

IEICE TRANSACTIONS on Information and Systems   Vol.E102-D   No.12   pp.2615-2618
Publication Date: 2019/12/01
Publicized: 2019/09/17
Online ISSN: 1745-1361
DOI: 10.1587/transinf.2019EDL8128
Type of Manuscript: LETTER
Category: Artificial Intelligence, Data Mining
artificial intelligence,  deep learning,  acoustic signal,  self-attention,  CNN,  

Full Text: PDF>>
Buy this Article

In-class species classification based on animal sounds is a highly challenging task even with the latest deep learning technique applied. The difficulty of distinguishing the species is further compounded when the number of species is large within the same class. This paper presents a novel approach for fine categorization of animal species based on their sounds by using pre-trained CNNs and a new self-attention module well-suited for acoustic signals The proposed method is shown effective as it achieves average species accuracy of 98.37% and the minimum species accuracy of 94.38%, the highest among the competing baselines, which include CNN's without self-attention and CNN's with CBAM, FAM, and CFAM but without pre-training.