Noise source detection and recognition for audio indexing

我们研究了一种基于随机方法的音频源检测方法，用于检测语音、噪声、音乐和静音。我们的方法不仅使用传统的表面声学特征，如信号能量和音调频率，但也基于频谱相关性更准确的检测新功能。广播新闻的实验表明，这些特征参数使得更准确地捕获音频源成为可能。本研究亦提出一种基于噪声模型的音源侦测方法。为了准确检测，我们设计了两种方法来生成多个噪声模型，通过聚类技术。一种方法是基于逐帧数据相似性，另一种是基于噪声源相似性。前一种方法采用K均值聚类和平滑技术，以避免不准确的分割。后一种方法涉及噪声建模的基础上产生的噪声集群的逐步合并的树数据结构。分类实验表明，通过使用这些方法，音频源可以检测到更好的准确性比传统的方法实现。当使用后一种方法产生的四个噪声模型时，对于声源不重叠的时段，噪声检测性能提高了3.9%。对于包括重叠段的音频流的实验，噪声检测性能提高了1.2%，而语音检测性能没有降低。

We have studied an audio source detection approach based on a stochastic method to detect speech, noise, music, and silence. Our approach uses not only conventional surface acoustic features such as signal energy and pitch frequency but also new features that are based on spectral correlation for more accurate detection. The experiment with the broadcast news demonstrated that these feature parameters made it possible to capture the audio source more accurately. This research also proposed a sound source detection approach based on elaborate noise-modeling techniques for audio indexing. For accurate detection, we devised two methods to generate multiple-noise models through clustering techniques. One method is based on frame-wise data similarity, and the other is based on noise source similarity. The former method employs K-means clustering and a smoothing technique to avoid inaccurate segmentation. The latter method involves noise modeling based on a tree data structure generated by the progressive merging of noise clusters. The classification experiments show that by using these proposed methods, audio sources can be detected with better accuracy than that achieved by the conventional methods. When four noise models generated by the latter method were used, the noise detection performance increased by 3.9% for the periods in which the sound sources did not overlap. With regard to the experiments for an audio stream that included overlapped segments, the noise detection performance increased by 1.2% without a decrease in the speech detection performance.