A Novel Scheme to classify Read and Spontaneous Speech
This addresses a domain-specific need for improving remote interview analysis, but it is incremental as it applies an existing method to a new task.
The paper tackles the problem of distinguishing between read and spontaneous speech in audio recordings, which is important for remote telephonic interviews during the COVID-19 pandemic, and achieves effective classification using a small set of features derived from a pre-trained DeepSpeech engine.
The COVID-19 pandemic has led to an increased use of remote telephonic interviews, making it important to distinguish between scripted and spontaneous speech in audio recordings. In this paper, we propose a novel scheme for identifying read and spontaneous speech. Our approach uses a pre-trained DeepSpeech audio-to-alphabet recognition engine to generate a sequence of alphabets from the audio. From these alphabets, we derive features that allow us to discriminate between read and spontaneous speech. Our experimental results show that even a small set of self-explanatory features can effectively classify the two types of speech very effectively.