arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2303.03005cs.SDcs.LGeess.AS
收藏

基于Conv-Tasnet的设备上低复杂度声源分离缩放策略

Scaling strategies for on-device low-complexity source separation with Conv-Tasnet

University of Trento(特伦托大学);Fondazione Bruno Kessler(布鲁诺·凯斯勒基金会)

Mohamed Nabih Ali, Francesco Paissan, Daniele Falavigna, Alessio Brutti

摘要:

最近,一些非常有效的单通道语音分离的神经网络方法已经在文献中提出。然而,由于这些模型的尺寸和复杂性,它们在低资源设备(例如助听器和耳机)上的使用仍然是一个挑战,并且尚未提供既定的解决方案。虽然已经提出了基于修剪或压缩神经模型的方法,但适合于某个应用领域的模型架构的设计通常需要启发式程序,这些程序不容易移植到不同的低资源平台。鉴于众所周知的Conv-Tasnet语音分离架构的模块化性质,在本文中,我们考虑直接控制模型的整体大小的三个参数,即:残差块的数量,分离块的重复次数和深度卷积中的通道数量,并通过实验评估它们如何影响语音分离性能。特别是,在Libri 2 Mix上进行的实验表明,膨胀的1D-Conv块的数量是最关键的参数,并且在残差块中使用额外膨胀可以减少性能下降。

英文摘要:

Recently, several very effective neural approaches for single-channel speech separation have been presented in the literature. However, due to the size and complexity of these models, their use on low-resource devices, e.g. for hearing aids, and earphones, is still a challenge and established solutions are not available yet. Although approaches based on either pruning or compressing neural models have been proposed, the design of a model architecture suitable for a certain application domain often requires heuristic procedures not easily portable to different low-resource platforms. Given the modular nature of the well-known Conv-Tasnet speech separation architecture, in this paper we consider three parameters that directly control the overall size of the model, namely: the number of residual blocks, the number of repetitions of the separation blocks and the number of channels in the depth-wise convolutions, and experimentally evaluate how they affect the speech separation performance. In particular, experiments carried out on the Libri2Mix show that the number of dilated 1D-Conv blocks is the most critical parameter and that the usage of extra-dilation in the residual blocks allows reducing the performance drop.

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。

↑