Low-Key Shallow Learning Voice Spoofing Detection System


Ali D., Al Shareeda S. Y. A., Abdulrahman N.

4th IEEE Middle East and North Africa COMMunications Conference, MENACOMM 2022, Amman, Jordan, 6 - 08 December 2022, pp.77-82 identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/menacomm57252.2022.9998199
  • City: Amman
  • Country: Jordan
  • Page Numbers: pp.77-82
  • Keywords: classic machine learning, Constant Q Cepstral Coefficients (CQCC), data security, GMM classifier, Melfrequency Cepstral Coefficients (MFCC), spoofed voice detection
  • Istanbul Technical University Affiliated: Yes

Abstract

This paper creates a Gaussian shallow learning Mixture Model (GMM) voice-replay detector using the MATLAB low-key machine learning and statistics libraries. Our model extracts the Mel frequency cepstrum coefficients (MFCC) and constant Q cepstrum coefficients (CQCC) from the input voice signal in the front-end feature extraction stage. The collected characteristics are fed to the constructed GMM classifier to categorize the input voice as either authentic from a live source or replayed from a prerecorded source. The GMM is trained using large datasets of voice feature samples representing both classes. The classifier's performance is measured using the Equal Error Rate (%EER) metric. To optimize performance, we subject the trained GMM to substantial development and assessment datasets in diverse scenarios and settings of reduction, normalization, and filtration. The best %EER results for the GMM classifier are 11.2237% for the development set and 22.5429% for the evaluation set.