Acoustic Signal Processing Revisited: Exploring Hilbert-Huang Transform and the Challenge of Cross-Term Errors

پذیرفته شده برای پوستر
کد مقاله : 1061-ISAV2024 (R1)
نویسندگان
1Harbin Institute of Technology
2Principal Acoustic Consultant, Hawkins & Associates
چکیده
Numerous comprehensive studies have investigated the effectiveness of joint time-frequency transformations for analyzing non-stationary time series data. The primary objective of these investigations is to improve accuracy in both time and frequency domains, which is crucial for a wide range of applications. These methods have demonstrated significant efficacy in various research fields, particularly in audio and acoustic signal processing. Despite their success, several challenges persist, such as the occurrence of cross-term errors. This paper presents a comparative analysis of two prominent time-frequency analysis methods: The Short-Time Fourier Transform (STFT) and the Hilbert-Huang Transform (HHT). We employ three acoustic signal types, drawn from industrial applications, music, and audio processing, to evaluate the performance of each method. Our findings reveal that the STFT outperforms the HHT, providing more accurate results across all tested signal types. Notably, the HHT introduces a higher risk of cross-term errors, which can compromise the clarity and usability of the analyzed data.
کلیدواژه ها
 
Title
Acoustic Signal Processing Revisited: Exploring Hilbert-Huang Transform and the Challenge of Cross-Term Errors
Authors
JAVAD ISAVAND, Jihong Yan, Andrew Peplow
Abstract
Numerous comprehensive studies have investigated the effectiveness of joint time-frequency transformations for analyzing non-stationary time series data. The primary objective of these investigations is to improve accuracy in both time and frequency domains, which is crucial for a wide range of applications. These methods have demonstrated significant efficacy in various research fields, particularly in audio and acoustic signal processing. Despite their success, several challenges persist, such as the occurrence of cross-term errors. This paper presents a comparative analysis of two prominent time-frequency analysis methods: The Short-Time Fourier Transform (STFT) and the Hilbert-Huang Transform (HHT). We employ three acoustic signal types, drawn from industrial applications, music, and audio processing, to evaluate the performance of each method. Our findings reveal that the STFT outperforms the HHT, providing more accurate results across all tested signal types. Notably, the HHT introduces a higher risk of cross-term errors, which can compromise the clarity and usability of the analyzed data.
Keywords
Acoustic Signal Processing, Hilbert-Huang transform, Cross-term Error