引导式测验 · 中等 等级
波束成形和麦克风阵列测验
查看在嘈杂的声学环境中影响定向音频捕获的阵列延迟、转向几何结构、MVDR 约束和间距限制。
问题 1 的 8
波束成形主要利用什么物理线索?
预览每个问题8 问题
当您参加测验时,就会出现答案和解释。
- 波束成形主要利用什么物理线索?
- 多个麦克风的到达时间/相位差
- 声源颜色
- 设备的电池电量
- 房间的温度
- 最简单的波束形成技术是什么?
- 傅里叶折叠
- 延迟求和
- 光谱门控
- 音高追踪
- What is the defining constraint in an MVDR beamformer?
- Maximize every incoming direction equally
- Remove the desired source before summation
- Preserve the desired target response while minimizing output power
- Convert the microphone signals into text
- 波束成形与单麦克风噪声抑制有何关系?
- 它完全取代了它并使其过时
- 和它是一模一样的
- 它只在没有任何其他处理的情况下起作用
- 它通过添加单个麦克风缺乏的空间信息来补充它
- Why can microphone spacing that is too large create spatial aliasing?
- Different directions can produce indistinguishable sampled phase differences at a frequency
- The microphones stop recording sound
- The sample rate automatically drops to zero
- The array loses all time information
- 什么是“神经波束形成”?
- 波束形成完全由手工完成
- 将深度网络(例如掩模或方向估计)与空间波束形成器相结合
- 卸下所有麦克风
- 一种单通道变调器
- What can happen when an MVDR beamformer’s assumed steering vector is inaccurate?
- The array automatically gains microphones
- The audio changes into a spectrogram image
- Its target preservation or interference suppression can degrade
- The desired source is always amplified without limit
- What information determines the steering delays for a chosen look direction?
- Audio file format and transcript length
- Speaker age and language
- 仅模型参数数量
- Target direction, microphone geometry, and speed of sound