Abstract:

A comprehensive method applying a nonlinear frequency compression (FC) as complementary to multiband loudness compensation is proposed, which is able to improve loudness compensation and simultaneously increase high-frequency speech intelligibility for digital hearing aids. First, to avoid the spectral distortions in the sounds, speech intelligibility-based frequency spectrum splitting is introduced. Then a nonlinear FC (NLFC) is also proposed to compress the high-frequency sounds to the lower bands where the audibility is available. Moreover, the introduced NLFC adjusts compression ratio (CR) based on the speech intelligibility percentage in different frequency ranges. Finally, an adaptive wide dynamic range compression (AWDRC) with a time-varying CR is applied to achieve adaptive loudness compensation, and prevent the stationary CR of the typical wide dynamic range compression (WDRC) from generating a negative impact on the speech. The experimental test results show that the mean speech identification is improved at least 20% points in comparison with the typical WDRC or the conventional FC.

Key words: digital hearing aid, loudness compensation, AWDRC, frequency compression

摘要：

为了改进数字助听器中的响度补偿, 并提高高频部分的语音可懂度, 提出一种基于非线性频率压缩的多通道响度补偿的综合方法。首先, 为了避免语音的频率畸变, 基于语音可懂度进行频谱的多通道划分。然后, 采用一种非线性的频率压缩方法, 将高频部分的声音压缩至患者能听到的低频部分。所提出的非线性频率压缩方法是基于不同频段对语音理解度的贡献占比来改变频率压缩比。最后, 为了实现自适应的响度补偿同时防止传统宽动态范围压缩的固定压缩比降低语音质量, 采用一种随时间可变压缩比的自适应宽动态范围压缩方法。实验结果表明, 相对于传统的宽动态范围压缩和频率压缩方法, 该方法可以改善20%的语音鉴别准确率。

关键词: 数字助听器, 响度补偿, 自适应宽动态范围压缩(AWDRC), 频率压缩