Spatial hearing augmentation and hearing devices

人类听觉系统能够通过利用来自两只耳朵的双耳线索，以及来自耳廓的角度依赖滤波的频谱线索，以空间方式定位声源。然而，助听器经常干扰这些线索。这是由于增强音频信号所需的动态压缩经常会干扰双耳信号，以及助听器的麦克风经常放置在耳道之外，从而消除了耳廓的任何频谱滤波。这是有问题的，因为研究表明，在有多个说话者的情况下，空间听力对语音清晰度至关重要，这通常被称为“鸡尾酒会”问题。本博士的目的是使用空间音频渲染技术来恢复这些空间线索，并通过调整/夸大这些线索来探索这些线索的增强，并评估这对听力受损个体的感知益处。

The human auditory system is able to locate sound sources in a spatial manner through the utilization of binaural cues from the two ears, as well as spectral cues from the angular dependent filtering of the pinnae. Hearing aids, however, often disturb these cues. This is due to the fact that the dynamic compression needed to enhance the audio signal often disturbs binaural cues, as well as the fact that the microphone of the hearing aids are often placed outside of the ear canal, thus removing any of the spectral filtering of the pinnae. This is problematic, as studies have shown that spatial hearing is essential to speech intelligibility in scenarios with multiple speakers, often dubbed the "Cocktail Party" problem. The aim of this PhD is to restore these spatial cues using spatial audio rendering techniques, as well as exploring the enhancement of these cues through the adjustment/exaggeration of these cues and evaluating where this provides a perceptual benefit to hearing impaired individuals.