基于图中心性的滤波器剪枝压缩音频CNN
Compressing audio CNNs with graph centrality based filter pruning
AI总结：
针对音频CNN存在冗余滤波器导致计算与内存开销高的问题，提出基于图中心性衡量滤波器共性并剪枝高共性滤波器的框架，在两类音频任务上实现效率提升且性能损失极小甚至略有提升。
摘要：
卷积神经网络（CNN）是音频分类等诸多现实问题高性能解决方案中的常用模型。CNN包含大量参数与滤波器，其中部分对性能的影响远大于其余部分，这意味着网络可能存在大量冗余滤波器，在提升性能作用有限的同时，增加了CNN的计算与内存开销。为提升CNN的效率，我们提出一种剪枝框架，可剔除“共性”最高的滤波器，我们采用图论中的“中心性”概念来衡量这种共性。我们假设高中心性的滤波器应被剔除，因为它代表共性特征，可被其他滤波器替代且不会对网络性能造成明显影响。我们在声学场景分类与音频标注任务上对所提框架开展了实验评估：在DCASE 2021 Task 1A基线网络上，相比原始网络，我们的方法将单次推理计算量降低71%，参数减少50%，而准确率下降不足2个百分点；针对PANNs这类面向音频标注设计的大规模CNN，我们的方法将单次推理计算量降低24%，参数减少41%，同时性能略有提升。
英文摘要：
Convolutional neural networks (CNNs) are commonplace in high-performing solutions to many real-world problems, such as audio classification. CNNs have many parameters and filters, with some having a larger impact on the performance than others. This means that networks may contain many unnecessary filters, increasing a CNN's computation and memory requirements while providing limited performance benefits. To make CNNs more efficient, we propose a pruning framework that eliminates filters with the highest "commonality". We measure this commonality using the graph-theoretic concept of "centrality". We hypothesise that a filter with a high centrality should be eliminated as it represents commonality and can be replaced by other filters without affecting the performance of a network much. An experimental evaluation of the proposed framework is performed on acoustic scene classification and audio tagging. On the DCASE 2021 Task 1A baseline network, our proposed method reduces computations per inference by 71\% with 50\% fewer parameters at less than a two percentage point drop in accuracy compared to the original network. For large-scale CNNs such as PANNs designed for audio tagging, our method reduces 24\% computations per inference with 41\% fewer parameters at a slight improvement in performance.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
