APSIPA Transactions on Signal and Information Processing > Vol 12 > Issue 3

Missing Data Completion of Multi-channel Signals Using Autoencoder for Acoustic Scene Classification

Yuki Shiroma, Tokyo Metropolitan University, Japan, shiroma-yuki@ed.tmu.ac.jp , Yuma Kinoshita, Tokai University, Japan, Keisuke Imoto, Doshisha University, Japan, Sayaka Shiota, Tokyo Metropolitan University, Japan, Nobutaka Ono, Tokyo Metropolitan University, Japan, Hitoshi Kiya, Tokyo Metropolitan University, Japan
 
Suggested Citation
Yuki Shiroma, Yuma Kinoshita, Keisuke Imoto, Sayaka Shiota, Nobutaka Ono and Hitoshi Kiya (2023), "Missing Data Completion of Multi-channel Signals Using Autoencoder for Acoustic Scene Classification", APSIPA Transactions on Signal and Information Processing: Vol. 12: No. 3, e16. http://dx.doi.org/10.1561/116.00000074

Publication Date: 25 Apr 2023
© 2023 Y. Shiroma, Y. Kinoshita, K. Imoto, S. Shiota, N. Ono and H. Kiya
 
Subjects
 
Keywords
Missing data completionAutoencoderMulti-channel signalsAcoustic scene classification
 

Share

Open Access

This is published under the terms of CC BY-NC.

Downloaded: 563 times

In this article:
Introduction 
Problems of Multi-channel Acoustic Scene Classification 
Missing Data Completion Using Autoencoder 
Experiment 
Conclusion 
References 

Abstract

In this paper, we propose an autoencoder-based missing data completion method for multi-channel acoustic scene classification (ASC). It has been reported that many deep-learning-based ASC methods using multi-channel signals have robust performance. The advantage of using multi-channel data is the capture of spatial and frequency information. However, when there is missing data in multi-channel signals, the classification performance declines significantly. We focus on completing the missing data by using an autoencoder as the preprocessor of ASC models. Since positional relationships between multi-channel microphones are modeled in the latent space of the proposed autoencoder, missing information is reconstructed via the latent space from the multi-channel input, including missing data. In an experiment, the missing data is completed by using the proposed autoencoder, and the accuracy of ASC systems is improved by using the completed multi-channel signals.

DOI:10.1561/116.00000074

Companion

APSIPA Transactions on Signal and Information Processing Special Issue - Advanced Acoustic, Sound and Audio Processing Techniques and Their Applications
See the other articles that are part of this special issue.