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 2304.07142cs.SDcs.AIcs.LGcs.NEeess.AS
收藏

神经网络语音分离模型训练的数据采样策略研究

On Data Sampling Strategies for Training Neural Network Speech Separation Models

The University of Sheffield(谢菲尔德大学)

William Ravenscroft, Stefan Goetze, Thomas Hain

AI总结:

本研究分析训练信号长度限制对SepFormer、Conv-TasNet两种语音分离模型的影响,发现特定长度限制可通过增加独特训练样本提升性能,4.42秒限制结合动态混合训练的SepFormer性能与无限制最优模型相当,且训练时间减少44%。

摘要:

语音分离仍是多说话人信号处理领域的重要研究方向。深度神经网络(DNN)模型已在众多语音分离基准上取得了最优性能,但部分模型训练耗时久、内存需求高。已有研究提出缩短训练样本以解决这些问题,但该操作对模型性能的影响尚未得到充分厘清。\n本研究针对两种语音分离模型——Transformer架构的SepFormer与卷积架构的Conv-TasNet,分析了施加训练信号长度(TSL)限制的影响。研究从信号长度分布及其对训练效率的影响角度,分析了WJS0-2Mix、WHAMR和Libri2Mix数据集。结果表明,针对特定的信号长度分布,施加特定的TSL限制可提升模型性能,这主要是因为随机采样波形起始索引能生成更多独特的训练样本。\n实验显示,采用4.42秒TSL限制结合动态混合(DM)训练的SepFormer模型,性能与采用DM且无信号长度限制训练的最优SepFormer模型相当。此外,在WHAMR数据集上,4.42秒的TSL限制可使训练时间减少44%。

英文摘要:

Speech separation remains an important area of multi-speaker signal processing. Deep neural network (DNN) models have attained the best performance on many speech separation benchmarks. Some of these models can take significant time to train and have high memory requirements. Previous work has proposed shortening training examples to address these issues but the impact of this on model performance is not yet well understood. In this work, the impact of applying these training signal length (TSL) limits is analysed for two speech separation models: SepFormer, a transformer model, and Conv-TasNet, a convolutional model. The WJS0-2Mix, WHAMR and Libri2Mix datasets are analysed in terms of signal length distribution and its impact on training efficiency. It is demonstrated that, for specific distributions, applying specific TSL limits results in better performance. This is shown to be mainly due to randomly sampling the start index of the waveforms resulting in more unique examples for training. A SepFormer model trained using a TSL limit of 4.42s and dynamic mixing (DM) is shown to match the best-performing SepFormer model trained with DM and unlimited signal lengths. Furthermore, the 4.42s TSL limit results in a 44% reduction in training time with WHAMR.

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。

↑