Objective evaluation method for VoIP speech

Abstract

The invention discloses an objective evaluation method for VoIP speech, which comprises the following steps of: acquiring original speech at a transmitting end of VoIP; acquiring distortion speech introducing damage at a receiving end; differentiating the original speech and the distortion speech through a pre-processing module to acquire statement information, calculating coder / decoder damage of aligned statements by using a coder / decoder damage assessment module A, calculating the quality damage of the stretched or compressed statements by using a coder / decoder damage assessment module B, and evaluating the influence of end-to-end delay and internal delay on the speech by using a delay damage assessment module; and finally acquiring an objective evaluation score of speech quality by using a fitting module. The method evaluates the speech quality by using a network coder / decoder, an adaptive jitter buffer and the like, is more accurate compared with the method only using perceptual evaluation of speech quality (PESQ), considers the delay damage, and also can evaluate the session quality. Meanwhile, compared with the correlation degree with a subjective score acquired by an E model, the method is more suitable for objectively evaluating the speech with network damage.

Description

technical field [0001] The invention relates to the field of evaluation methods for communication network quality, in particular to an objective evaluation method for VoIP voice. Background technique [0002] There are currently two testing methods for voice quality: the traditional subjective evaluation method and the new objective evaluation method. Subjective evaluation is to evaluate speech and speech quality with people as the main body. Among the subjective evaluation methods of voice quality, MOS (Mean Opinion Score: Mean Option Score) is the most widely used score. It uses the average opinion score to measure the voice quality, and uses five grades to represent the voice quality level: excellent (5 points) , Good (4 points), Fair (3 points), Poor (2 points), Bad (1 point). Subjective listening test methods include: 1). ACR (Absolute Category Rating): absolute rating, which is the most commonly used listening test method based on sentence groups, sentence groups are...

AI Technical Summary

Problems solved by technology

4).QRDM (Quanta-Response Detectability Method): Quantitative response detectability method, used for quantitative analysis; 5).TM (Threshold Method): Threshold method. Time-consuming, labor-intensive and expensive, with insufficient flexibility, poor repeatability and stability, and relatively subject to human subjective influence, etc.

However, the E-model has the following disadvantages: firstly, it has not been verified enough in bounds and adequately measured in the laboratory for a large number of possible input parameter combinations, and the estimation is still in the stage of doubt and research; secondly, it simply It is considered that individual quality impairment factors (such as mute echo delay and distortion) are independent of each other, and they are all measured by the same scale; again, the premise is to assume that voice quality impairment factors are always physically attached, but studies have shown that this assumption is Reliability is questionable in some cases; in addition, it is based on the 300Hz-3400Hz telephone voice bandwidth, and cannot be accurately used for wideband voice quality assessment, etc.

Therefore, the currently widely used speech objective evaluation algorithm is perceptual speech quality evaluation - PESQ, but it cannot evaluate the speech that has passed through the network codec and adaptive jitter buffer, and the experiment proves that the main Very low objective correlation

Images