| 杨雪同,夏秀渝.基于注意力的双层级并行声学场景分类方法[J].,2025,43(3):588-595 |
| 基于注意力的双层级并行声学场景分类方法 |
| Attention-based dual-hierarchy parallel acoustic scene classification method |
| 投稿时间：2024-01-03 修订日期：2025-04-25 |
| 中文摘要: |
| 声学场景分类是计算机听觉任务之一，其通过对声频信号的分析，将声频分类为特定的场景类型。该技术可广泛应用于智能设备、声频监控等领域。声学场景自上而下可分为高层级场景，再细分为低层级场景。与直接针对低层级场景分类的方法不同，根据该层级关系提出一种基于注意力的双层级并行网络用于声学场景分类。首先基于残差网络构建并行的高低层级声学场景分类模型，从高层级分类模型间层特征中获取全局参考特征。然后根据全局参考特征和低层级分类模型特征间距离计算注意力权重，使低层级分类模型关注重要特征。最后利用增强推理层融合高低层级分类模型的输出。并行网络在DCASE2019任务1数据集上的准确率为89.5%，应用增强推理层后的准确率为90.1%，验证了所提网络模型和增强推理层的有效性。 |
| 英文摘要: |
| Acoustic scene classification is one of the computer auditory tasks, which classifies audio into specific scene types through the analysis of audio signals. This technology can be widely applied in fields such as smart devices and audio monitoring. The acoustic scene can be divided into high-level scenes and then subdivided into low-level scenes. Unlike methods that directly target low-level scene classification, an attention-based dual-hierarchy parallel network is proposed for acoustic scene classification based on the hierarchical relationship. Firstly, a parallel high-low level acoustic scene classification model is constructed utilizing residual networks, and global reference features are extracted from the intermediate features of the high-level classification model. Then, attention weights are computed by considering the distance between the global reference features and the low-level classification model features, this allows the low-level classification model to prioritize significant features. Finally, an enhanced inference layer is employed to integrate the output of high-low level classification models. The accuracy of this parallel network on the DCASE2019 Task 1 dataset is 89.5%, and the accuracy after applying the enhanced inference layer is 90.1%, verifying the effectiveness of the proposed network model and the enhanced inference layer. |
| DOI：10.11684/j.issn.1000-310X.2025.03.007 |
| 中文关键词: 声学场景分类 残差网络 注意力 层级关系 增强推理 |
| 英文关键词: Acoustic scene classification Residual network Attention Hierarchy Enhance inference |
| 基金项目: |
|
| 摘要点击次数: 11 |
| 全文下载次数: 4 |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
