Arbitrary Discrete Fourier Analysis and Its Application in Replayed Speech Detection

2 Mar 2024  ·  Shih-kuang Lee ·

In this paper, a group of finite sequences and its variants were proposed to use in conducting signal analysis; we called the developed signal analysis methods arbitrary discrete Fourier analysis (ADFA), Mel-scale discrete Fourier analysis (MDFA) and constant Q analysis (CQA). The effectiveness of three signal analysis methods were then validated by testing their performance on a replayed speech detection benchmark (i.e., the ASVspoof 2019 Physical Access) along with a state-of-the-art model. Comparable performance to the best reported systems were shown by the experimental results with three signal analysis methods. Furthermore, the CQA method shown its efficiency with less computation time in compared to the convention method constant Q transform (CQT), which is commonly used in spoofed and fake speech detection and music processing.

PDF Abstract

Datasets


  Add Datasets introduced or used in this paper

Results from the Paper


  Submit results from this paper to get state-of-the-art GitHub badges and help the community compare results to other papers.

Methods


No methods listed for this paper. Add relevant methods here