Oscillating Statistical Moments for Speech Polarity Detection
This addresses the need for automatic polarity detection to ensure reliable speech processing, but it is incremental as it builds on existing methods with a novel twist.
The paper tackled the problem of speech polarity detection, which is crucial for speech processing performance, by proposing a method based on oscillating statistical moments that exploit phase shifts dependent on polarity, achieving substantial improvement over state-of-the-art techniques on 10 speech corpora.
An inversion of the speech polarity may have a dramatic detrimental effect on the performance of various techniques of speech processing. An automatic method for determining the speech polarity (which is dependent upon the recording setup) is thus required as a preliminary step for ensuring the well-behaviour of such techniques. This paper proposes a new approach of polarity detection relying on oscillating statistical moments. These moments have the property to oscillate at the local fundamental frequency and to exhibit a phase shift which depends on the speech polarity. This dependency stems from the introduction of non-linearity or higher-order statistics in the moment calculation. The resulting method is shown on 10 speech corpora to provide a substantial improvement compared to state-of-the-art techniques.