DEW:一种波束方法来检测罕见的声音事件
Sania Gul1,2, Muhammad Salman Khan3, Ata Ur-Rehman4
1Department of Electrical Engineering, University of Engineering and Technology, Peshawar, Pakistan.
PloS one
|March 28, 2024
概括
一个新的声音事件检测 (SED) 系统通过波形分析和机器学习有效地识别罕见事件. 这种轻量级的方法在有限的设备上提供实时性能,优于深度学习模型.
科学领域:
- 信号处理 信号处理
- 机器学习 机器学习
- 声学 声学 在声学上
背景情况:
- 在开放环境中检测罕见的声音事件,由于背景噪音和事件频率不高,因此存在重大挑战.
- 现有的声音事件检测 (SED) 深度学习模型通常需要大量的训练数据和计算资源.
- 需要高效和轻量级的SED系统,适合在资源有限的设备上实时应用.
研究的目的:
- 开发一种新且计算效率高的声音事件检测 (SED) 系统,用于识别罕见的声学事件.
- 对已建立的深度学习架构进行系统性能评估,重点关注准确性,效率和资源要求.
主要方法:
- 波段多分辨率分析 (MRA) 将音频信号分解,然后进行波段消音以减少背景噪声.
- K-medoids聚类和峰值查找算法识别出显著的过渡,表明罕见事件的发生.
- 波纹散射网络 (WSN) 从选定的音频段中提取特征,由支持矢量机器 (SVM) 分类.
主要成果:
- 拟议的SED框架实现了与基于卷积神经网络 (CNN) 的系统可比的错误率.
- 该算法表现出显著的计算效率,需要更少的训练时间和更少的数据块进行分析.
- 该系统在没有可学习的参数的情况下运行,不需要广泛的训练时代和数据增强.
结论:
- 开发的SED系统为深度学习模型提供了一种轻量级和计算效率高的替代方案,用于检测罕见的声音事件.
- 它处理较少数据段的能力和最小的培训要求使其非常适合边缘设备上的实时应用.
- 拟议的方法为在计算资源有限的开放环境声学监控提供了强大的解决方案.
相关概念视频
01:01Sound Waves
9.1K
Sound waves can be thought of as fluctuations in the pressure of a medium through which they propagate. Since the pressure also makes the medium's particles vibrate along its direction of motion, the waves can be modeled as the displacement of the medium's particles from their mean position.
Sound waves are longitudinal in most fluids because fluids cannot sustain any lateral pressure. In solids, however, shear forces help in propagating the disturbance in the lateral direction as well....
Sound waves are longitudinal in most fluids because fluids cannot sustain any lateral pressure. In solids, however, shear forces help in propagating the disturbance in the lateral direction as well....
9.1K
00:53Sound Waves: Interference
3.7K
Sound waves can be modeled either as longitudinal waves, wherein the molecules of the medium oscillate around an equilibrium position, or as pressure waves. When two identical waves from the same source superimpose on each other, the combination of two crests or two troughs results in amplitude reinforcement known as constructive interference. If two identical waves, that are initially in phase, become out of phase because of different path lengths, the combination of crests with troughs...
3.7K
01:31Hearing
52.1K
When we hear a sound, our nervous system is detecting sound waves—pressure waves of mechanical energy traveling through a medium. The frequency of the wave is perceived as pitch, while the amplitude is perceived as loudness.
52.1K
01:01Perception of Sound Waves
4.5K
The human ear is not equally sensitive to all frequencies in the audible range. It may perceive sound waves with the same pressure but different frequencies as having different loudness. Moreover, the perception of sound waves depends on the health of an individual's ears, which decays with age. The health of one's ears may also be affected by regular exposure to loud noises.
The pitch of a sound depends on the frequency and the pressure amplitude of the source. Two sounds of the same...
The pitch of a sound depends on the frequency and the pressure amplitude of the source. Two sounds of the same...
4.5K
01:16Shock Waves
2.0K
While deriving the Doppler formula for the observed frequency of a sound wave, it is assumed that the speed of sound in the medium is greater than the source's speed through it. When this condition is breached, a shock wave occurs.
When the source's speed approaches the speed of sound, constructive interference between successive wavefronts emitted by the source occurs immediately behind it. Initially, scientists believed that this constructive interference would result in such high...
When the source's speed approaches the speed of sound, constructive interference between successive wavefronts emitted by the source occurs immediately behind it. Initially, scientists believed that this constructive interference would result in such high...
2.0K
01:30Classification of Signals
456
In signal processing, signals are classified based on various characteristics: continuous-time versus discrete-time, periodic versus aperiodic, analog versus digital, and causal versus noncausal. Each category highlights distinct properties crucial for understanding and manipulating signals.
A continuous-time signal holds a value at every instant in time, representing information seamlessly. In contrast, a discrete-time signal holds values only at specific moments, often denoted as x(n), where...
A continuous-time signal holds a value at every instant in time, representing information seamlessly. In contrast, a discrete-time signal holds values only at specific moments, often denoted as x(n), where...
456