Neural Encoding Detection is Not All You Need for Synthetic Speech Detection

TL;DR

This paper reviews current synthetic speech detection methods, emphasizing the limitations of relying solely on neural encoding detection and guiding future research directions.

Contribution

It critically analyzes the focus on neural encoding detection and offers recommendations to diversify research approaches in synthetic speech detection.

Findings

Neural encoding detection alone may not be sufficient for robust synthetic speech detection

Current trends may overemphasize neural encoding, risking future obsolescence

Guidelines for future research directions in the field

Abstract

This paper reviews the current state and emerging trends in synthetic speech detection. It outlines the main data-driven approaches, discusses the advantages and drawbacks of focusing future research solely on neural encoding detection, and offers recommendations for promising research directions. Unlike works that introduce new detection methods or datasets, this paper aims to guide future state-of-the-art research in the field and to highlight the risk of overcommitting to approaches that may not stand the test of time.