Audio quality evaluating method and device, electronic device and storage medium

Abstract

The invention provides an audio quality evaluating method and device, an electronic device and a storage medium. The method includes the steps of obtaining the speed value of a sound maker corresponding to each voice segment according to the effective voice duration of each voice segment corresponding to an audio file and the corpora text corresponding to each voice segment, conducting statisticsanalysis according to the speed value of the sound maker corresponding to each voice segment and a preset rule to obtain a statistics result, and obtaining the quality evaluating result of the audio file according to the statistics result and preset conditions. By means of the method, engineering automatic analysis is conducted according to the speed reset rule of the sound maker, the audio quality evaluating efficiency can be effectively improved, and the coverage of detection samples is effectively improved.

Description

technical field [0001] The invention relates to the technical field of computer processing, in particular to an audio quality evaluation method, device, electronic equipment and storage medium. Background technique [0002] In the process of making the long free dialogue corpus, it is necessary to mark the start time and end time of each speech segment of each speaker in the long audio file, and also mark the corpus text corresponding to the voice in the speech segment. In the actual production process, there are often various problems that lead to low audio quality, for example: there are a large number of invalid silent segments in the voice segment, and in severe cases, there is even no valid voice content in the voice segment at all, or , the annotated corpus text of a speech segment does not match the actual audio content, for example, compared with the actual audio content of the speech segment, the marked corpus text is obviously missing, etc. Therefore, it is necess...

AI Technical Summary

Problems solved by technology

However, the above method needs to be done manually, which is time-consuming and labor-intensive, resulting in low efficiency of audio quality assessment, and the way of manual spot check is that the listener samples in multiple speech segments, and the number of samples drawn is limited, resulting in the coverage of the test samples. lower

Images