Spatial Audio for Managing Mental Illness (SAMMI)

人类的听觉系统是惊人的。不用思考，我们就知道声音从哪里来，我们可以有选择地听到一个声音，而另一个声音则在另一个声音面前。标准麦克风，比如手机或笔记本电脑中的麦克风，无法做到这一点。相反，他们把世界缩小到一个单一的频道。当大脑加班加点地工作，根据微妙的时间和上下文线索分离声音时，混乱就占了上风。该项目将由Quaudio的共同发明者大卫纽金特博士承担，Quaudio是一种能像我们一样听到世界的技术。大卫还与神经科学家合作，使用语音和音频分析诊断认知能力下降。Quaudio执行盲源分离：在不事先知道声源存在、数量或位置的情况下检测、跟踪和隔离声源。我们使用四个麦克风的紧凑阵列来实现这一点。这就是Quaudio的名字来源。SAMMI将在两个流中进行。在流1中，（四通道）麦克风将捕捉和解析空间声音，然后将这些声音通过编写到Windows应用程序中的代码进行处理和呈现。结果将是一个程序，可供精神卫生专业人员，患者和护理人员试用。同时，Stream 2将把Quaudio硬件整合到沉浸式设备中，如混合现实眼镜。将应用Strand 1中开发的代码，并测试跟踪和隔离声源的能力。其结果将是如何在虚拟和混合现实耳机中捕获、操纵和呈现空间音频的蓝图。越来越多的证据表明，空间音频在精神疾病的诊断、治疗和管理中起着至关重要的作用。我们将研究三个：1：广泛性焦虑症：如何通过双耳耳机捕获和控制空间音频及其呈现来减少焦虑？第二章：自闭症中的听觉超敏反应：应用机器学习自动衰减响亮的声源，而不会衰减所有声音或影响声学清晰度。3：痴呆症：通过双耳渲染和耳机自动将声源组织成均匀间隔的方向，以呈现声学场景。在所有三种情况下，我们都将与心理健康专业人士和领先的慈善机构展示和讨论我们的工作。他们的反馈将纳入我们的报告。

The human auditory system is amazing. Without thinking we know where sounds are coming from, and we can selectively listen to one voice in the presence of another.Standard microphones, such as the ones found in your phone or laptop, cannot do this. Instead, they reduce the world to a single channel. Confusion reigns, as the brain works overtime to isolate sounds based on subtle temporal and contextual clues.The project will be undertaken by Dr David Nugent, coinventor of Quaudio, a technology that hears the world as we do. David has also worked with neuroscientists on the diagnosis of cognitive decline using speech and audio analysis.Quaudio performs Blind Source Separation: the detection, tracking and isolation of sound sources with no prior knowledge of their presence, number or location. We do this using a compact array of four microphones. That is where Quaudio gets its name from.SAMMI will be undertaken in two Streams.In Stream 1, an ambisonic (four channel) microphone will capture and digitize spatial sounds, these will then be processed and rendered by code written into a Windows App. The result will be a program ready for trialing by mental health professionals, patients and caregivers alike.In parallel, Stream 2 will see the incorporation of Quaudio hardware into immersive devices such as mixed reality glasses. Code developed in Strand 1 will be applied, and the ability to track and isolate sound sources tested. The result will be a blueprint for how spatial audio can be captured, manipulated and presented in virtual and mixed reality headsets.There is a growing body of evidence the spatial audio plays a crucial role in the diagnosis, treatment and management of mental illness. We will examine three:1: Generalised anxiety disorder: How can the capture and control of spatial audio and its presentation via binaural headphones reduce anxiety?2: Auditory hypersensitivity in autism: Applying machine learning to automatically attenuate loud sound sources without attenuating everything or compromising acoustic clarity.3: Dementia: Declutter the acoustic scene by automatically organizing sound sources into evenly spaced directions for presentation via binaural rendering and headphones.In all three cases we will present and discuss our work with mental health professionals and leading charities. Their feedback will be included in our report.