A new objective speech quality evaluation method based on weighted Bark spectral distortion (WBSD) is presented.
提出了一种基于加权巴克谱失真(WBSD)的语音质量客观评价算法。
A novel approach to the real-time speech quality evaluation in a packet network using feed-forward multiple class random neural network (FFMCRNN) was presented.
提出了一种利用前馈随机神经网络在分组网络中进行实时语音质量评价的新方法。
In the application of speech quality objective assessment, the result of subjective evaluation under plenty of distortion conditions is directly used as the desired value of training.
结合语音质量客观评价应用，我们以在大量的失真条件下得到的主观评价结果作为期望值对该网络进行训练。
This paper studies briefly a distance measure for objective quality evaluation of speech communication channels using instantaneous and dynamic spectral features.
主要讨论了一种使用语音频谱动态特性的倒谱距离测量对通信系统或通信设备进行客观音质评价方法。
Testing with the average segment SNR and Perceptual Evaluation of Speech Quality (PESQ), the improved algorithm is close to g.
用平均分段信噪比和感知语音质量评价(PESQ)测试，改进算法编码质量接近于g。
Testing with the average segment SNR and Perceptual Evaluation of Speech Quality (PESQ), the improved algorithm is close to g.
用平均分段信噪比和感知语音质量评价(PESQ)测试，改进算法编码质量接近于g。
应用推荐