湘潭大学物理与光电工程学院，湘潭 411105
国家自然科学基金(62071411)资助项目；湖南省自然科学基金(2018JJ3486)资助项目。
School of Physics and Optoelectronics, Xiang Tan University, Xiangtan 411105, China
- 摘要 |
- 图/表 |
- 访问统计 |
- 参考文献 |
- 相似文献 |
- 引证文献 |
为了对环境声音进行更好的识别和分类，提出了基于多级残差网络（Multilevel residual network， Mul-EnvResNet）的环境声音分类方法。对声音事件进行时标和基频压扩之后，提取其梅尔频率倒谱系数（Mel-frequency cepstral coefficients， MFCCs），以及它们的差分作为特征参数送入Mul-EnvResNet对声音事件进行分类。实验数据集采用ESC-50，将Mul-EnvResNet模型与端到端的卷积神经网络（EnvNet）、基于注意力机制的循环神经网络（Attention based convolutional recurrent neural network， ACRNN），以及受限卷积玻尔兹曼机的无监督滤波器组模型（Convolutional restricted Boltzmann machine， ConvRBM）进行对比实验。实验结果表明， Mul-EnvResNet取得了89.32%的最佳分类准确率，相较上述3种模型在分类准确率上分别有18.32%、3.22%、2.82%的提升，相较于其他的声音分类方法也均有明显的优势。
To better identify and classify environmental sound， a multilevel residual network （Mul-EnvResNet） is proposed for environmental sound classification. After time stretch and pitch shift for sound events， the Mel-frequency cepstral coefficients （MFCCs） and their deltas are extracted as feature parameters and sent into the Mul-EnvResNet to classify sound events. The experimental data set uses ESC-50， Mul-EnvResNet is compared with the end-to-end convolutional neural network （EnvNet）， the attention based convolutional recurrent neural network （ACRNN） and the unsupervised filterbank learning using convolutional restricted Boltzmann machine （ConvRBM）. The experimental results show that， Mul-EnvResNet achieves the best accuracy rate of 89.32% in terms of classification accuracy， compared with the above three models， the classification accuracy has been improved by 18.32%， 3.22% and 2.82%， respectively， which also has obvious advantages compared with other sound classification methods.
引用本文
曾金芳,李友明,杨恢先,张钰,胡雅欣.基于多级残差网络的环境声音分类方法[J].数据采集与处理,2021,36(5):960-968
复制分享
文章指标
- 点击次数:
- 下载次数:
历史
- 收稿日期:2020-11-25
- 最后修改日期:2021-02-28
- 在线发布日期: 2021-09-25
