Detection of overlapping acoustic events based on NMF with shared basis vectors
Detection of overlapping acoustic events based on NMF with shared basis vectors
复制标题
DOI：
10.1109/gcce.2017.8229482
复制
发表时间：
2017-10
期刊：
影响因子：
--
通讯作者：
Kazumasa Yamamoto;Chikara Ishikawa;Koya Sahashi;S. Nakagawa
中科院分区：
文献类型：
--
作者：
Kazumasa Yamamoto;Chikara Ishikawa;Koya Sahashi;S. Nakagawa
声事件检测在计算声场景分析中起着重要的作用。虽然在真实的情况下会遇到合理的重叠问题，但传统的方法对重叠问题考虑不够。在本文中，我们提出了一种新的重叠声事件检测技术相结合的源分离技术的非负矩阵分解与共享基向量和基于深度神经网络的声学模型，以提高检测性能。我们的方法在基于帧的F-测量上显示出比D-CASE 2012挑战中取得的最佳结果高出20.0%的绝对性能。
Acoustic Event Detection plays an important role for computational acoustic scene analysis. Although we would face with a sound overlapping problem in a real situation, conventional methods do not consider the problem enough. In this paper, we propose a new overlapped acoustic event detection technique combined a source separation technique of Non-negative Matrix Factorization with shared basis vectors and a deep neural network based acoustic model to improve the detection performance. Our approach showed 20.0% absolute higher performance than the best result achieved in the D-CASE 2012 challenge on the frame based F-measure.