Artificial speech detection using image-based features and random forest classifier

Abstract

The ASVspoof 2015 Challenge was one of the efforts of the research community in the field of speech processing to foster the development of generalized countermeasures against spoofing attacks. However, most countermeasures submitted to the ASVspoof 2015 Challenge failed to detect the S10 attack effectively, the only attack that was generated using the waveform concatenation approach. Hence, more informative features are needed to detect previously unseen spoofing attacks. This paper presents an approach that uses data transformation techniques to engineer image-based features together with random forest classifier to detect artificial speech. The objectives are two-fold: (i) to extract image-based features from the melfrequency cepstral coefficients representation of the speech signal and (ii) to compare the performance of using the extracted features and Random Forest to determine the authenticity of voices with the existing approaches. An audio-to-image transformation technique was used to engineer new features in classifying genuine and spoof voices. An experiment was conducted to find the appropriate combination of the engineered features and classifier. Experimental results showed that the proposed approach was able to detect speech synthesis and voice conversion attacks effectively, with an equal error rate of 0.10% and accuracy of 99.93%.

A Comparative Evaluation on Data Transformation Approach for Artificial Speech Detection

“…Artificial speech detection techniques have witnessed significant evolution, falling into three primary categories: classic machine learning, end-to-end learning, and hybrid methodologies. Classic machine learning entails the manual crafting and extraction of predetermined features from data samples, which are then subject to separate classification modules [4,5]. In contrast, end-to-end learning orchestrates the automatic and joint identification and learning of all data sample features to determine their class labels.…”

“…RF was considered in this evaluation due to its demonstrated strong performance in the ASVspoof 2015 challenge [4]. RF is an ensemble learning model that employs decision trees for classification and regression.…”

Alzheimer image registration using hybrid random forest and deep regression network algorithm

“…Tan et al [19], showed a method for engineer image-based features when used with a RF classifier to identify artificial speech by use of data transformation techniques. The two goals are as follows: i) from the mel-frequency cepstral coefficients representation of the speech signal, extract image-based features; and ii) compare the effectiveness of using RF and the extracted features with the existing approaches to determine the authenticity of voices.…”