Abstract
Recently developed methods in spontaneous speech analytics require the use of speaker separation based on audio data, referred to as diarization. It is applied to widespread use cases, such as meeting transcription based on recordings from distant microphones and the extraction of the target speaker’s voice profiles from noisy audio. However, speech recognition and analysis can be hindered by background and point-source noise, overlapping speech, and reverberation, which all affect diarization quality in conjunction with each other. To compensate for the impact of these factors, there are a variety of supportive speech analytics methods, such as quality assessments in terms of SNR and RT60 reverberation time metrics, overlapping speech detection, instant speaker number estimation, etc. The improvements in speaker verification methods have benefits in the area of speaker separation as well. This paper introduces several approaches aimed towards improving diarization system quality. The presented experimental results demonstrate the possibility of refining initial speaker labels from neural-based VAD data by means of fusion with labels from quality estimation models, overlapping speech detectors, and speaker number estimation models, which contain CNN and LSTM modules. Such fusing approaches allow us to significantly decrease DER values compared to standalone VAD methods. Cases of ideal VAD labeling are utilized to show the positive impact of ResNet-101 neural networks on diarization quality in comparison with basic x-vectors and ECAPA-TDNN architectures trained on 8 kHz data. Moreover, this paper highlights the advantage of spectral clustering over other clustering methods applied to diarization. The overall quality of diarization is improved at all stages of the pipeline, and the combination of various speech analytics methods makes a significant contribution to the improvement of diarization quality.
Highlights
The widespread availability of tools for sound acquisition, as well as the cost reduction of audio data storage systems, require new methods for automatic processing
To deal with the inherent problems of multi-dialogue recordings such as speaker interruptions and the simultaneous utterances of multiple speakers, we examine options for fusing voice activity detector (VAD) with other speech analytic systems
We apply an automatic quality estimation (QE) system, described in [9], to cluster estimated speech-to-noise ratio (SNR)-RT60 vectors into speech and non-speech clusters and thereby retrieve an approximate voice activity markup. This method is not accurate, we investigate its usefulness in fusing it with the base deep neural network (DNN)-VAD
Summary
The widespread availability of tools for sound acquisition, as well as the cost reduction of audio data storage systems, require new methods for automatic processing Such tasks as the generation of meeting minutes, the processing of telephone conversations, and the automatic transcription of news or entertainment programs, require speech recognition [1], and involve audio annotation by speakers, which is usually referred to as speaker diarization. Preliminary information can influence the choice of diarization methods: whether the exact number of speakers is known or whether individual speech samples of their voices are available When it comes to meeting minutes, it is helpful to know in advance if the participants can move around the room or tend to interrupt each other. All of these factors can significantly affect the quality of diarization [3]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
