具有可学习间距的空洞卷积音频分类
Audio classification with Dilated Convolution with Learnable Spacings
AI总结：
本文将可学习间距空洞卷积（DCLS）应用于ConvNeXt、ConvFormer和FastViT架构，在不增加参数量且吞吐量成本极低的情况下，显著提升了AudioSet基准上的音频分类平均精度均值（mAP）。
摘要：
可学习间距空洞卷积（DCLS）是一种新型卷积方法，其卷积核元素位置在训练中通过反向传播学习。其在计算机视觉（ImageNet分类及下游任务）中的优势已获验证。本文证明DCLS同样适用于基于AudioSet分类基准的音频标签预测。我们选取了两种采用深度可分离卷积（DSC）的最先进架构ConvNeXt和ConvFormer，以及额外结合注意力机制的混合架构FastViT，将所有DSC层直接替换为DCLS层。这在未增加参数量且吞吐量成本极低的情况下，显著提升了这三种架构的平均精度均值（mAP）。基于PyTorch的方法代码发布于 https://github.com/K-H-Ismail/DCLS-Audio。
英文摘要：
Dilated convolution with learnable spacings (DCLS) is a recent convolution method in which the positions of the kernel elements are learned throughout training by backpropagation. Its interest has recently been demonstrated in computer vision (ImageNet classification and downstream tasks). Here we show that DCLS is also useful for audio tagging using the AudioSet classification benchmark. We took two state-of-the-art convolutional architectures using depthwise separable convolutions (DSC), ConvNeXt and ConvFormer, and a hybrid one using attention in addition, FastViT, and drop-in replaced all the DSC layers by DCLS ones. This significantly improved the mean average precision (mAP) with the three architectures without increasing the number of parameters and with only a low cost on the throughput. The method code is based on PyTorch and is available at https://github.com/K-H-Ismail/DCLS-Audio
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。
