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 1806.03185cs.SDeess.ASstat.ML
收藏

Wave-U-Net:一种用于端到端音频声源分离的多尺度神经网络

Wave-U-Net: A Multi-Scale Neural Network for End-to-End Audio Source Separation

Queen Mary University of London(伦敦玛丽女王大学);Spotify

Daniel Stoller, Sebastian Ewert, Simon Dixon

摘要:

用于音频声源分离的模型通常在幅度谱上操作,这忽略了相位信息,并且使得分离性能依赖于谱前端的超参数。因此,我们调查端到端的声源分离在时域,它允许建模相位信息,并避免固定的频谱变换。由于音频的高采样率,在采样级上采用长时间输入上下文是困难的,但是由于长范围时间相关性而需要高质量的分离结果。在这种情况下,我们提出了Wave-U-Net,这是U-Net对一维时域的适应,它重复地对特征图进行重新采样,以计算和组合不同时间尺度的特征。我们引入了进一步的架构改进,包括一个输出层,强制执行源的可加性,上采样技术和上下文感知的预测框架,以减少输出文物。歌声分离的实验表明,我们的架构产生的性能相媲美的最先进的基于频谱图的U-Net架构,给定相同的数据。最后,我们揭示了一个问题,离群值在目前使用的SDR评估指标,并建议报告排名为基础的统计数据,以减轻这个问题。

英文摘要:

Models for audio source separation usually operate on the magnitude spectrum, which ignores phase information and makes separation performance dependant on hyper-parameters for the spectral front-end. Therefore, we investigate end-to-end source separation in the time-domain, which allows modelling phase information and avoids fixed spectral transformations. Due to high sampling rates for audio, employing a long temporal input context on the sample level is difficult, but required for high quality separation results because of long-range temporal correlations. In this context, we propose the Wave-U-Net, an adaptation of the U-Net to the one-dimensional time domain, which repeatedly resamples feature maps to compute and combine features at different time scales. We introduce further architectural improvements, including an output layer that enforces source additivity, an upsampling technique and a context-aware prediction framework to reduce output artifacts. Experiments for singing voice separation indicate that our architecture yields a performance comparable to a state-of-the-art spectrogram-based U-Net architecture, given the same data. Finally, we reveal a problem with outliers in the currently used SDR evaluation metrics and suggest reporting rank-based statistics to alleviate this problem.

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。

↑