CWS-PResUNet:基于逐行子带相位感知的音乐声源分离ResUNet
CWS-PResUNet: Music Source Separation with Channel-wise Subband Phase-aware ResUNet
摘要:
音乐声源分离(MSS)近年来在深度学习模型方面取得了积极进展。许多MSS模型通过估计有界比率掩模和重用混合物的相位来对谱图进行分离。当使用卷积神经网络(CNN)时,权重通常在卷积期间在频谱图内共享,而不管频带之间的不同模式。在这项研究中,我们提出了一个新的MSS模型,通道的子带相位感知ResUNet(CWS-PResUNet),信号分解成子带,并估计一个未绑定的复杂的理想比率掩模(crescent)为每个源。CWS-PRESUNet利用信道子带(CWS)特征来限制频谱图上不必要的全局权重共享并减少计算资源消耗。节省的计算成本和存储器可以反过来允许更大的架构。在MUSDB 18 HQ测试集上,我们提出了一个276层的CWS-PRESNet,并以8.92的信号失真比(SDR)得分在人声上实现了最先进的(SoTA)性能。通过结合CWS-PRESUNet和Demucs,我们的ByteMSS系统在2021年ISMIR音乐分层(MDX)挑战赛有限训练数据赛道(排行榜A)中获得人声评分第2名和平均评分第5名。我们的代码和预训练模型可在以下网址公开获取:https://github.com/haoheliu/2021-ISMIR-MSS-Challenge-CWS-PResUNet
英文摘要:
Music source separation (MSS) shows active progress with deep learning models in recent years. Many MSS models perform separations on spectrograms by estimating bounded ratio masks and reusing the phases of the mixture. When using convolutional neural networks (CNN), weights are usually shared within a spectrogram during convolution regardless of the different patterns between frequency bands. In this study, we propose a new MSS model, channel-wise subband phase-aware ResUNet (CWS-PResUNet), to decompose signals into subbands and estimate an unbound complex ideal ratio mask (cIRM) for each source. CWS-PResUNet utilizes a channel-wise subband (CWS) feature to limit unnecessary global weights sharing on the spectrogram and reduce computational resource consumptions. The saved computational cost and memory can in turn allow for a larger architecture. On the MUSDB18HQ test set, we propose a 276-layer CWS-PResUNet and achieve state-of-the-art (SoTA) performance on vocals with an 8.92 signal-to-distortion ratio (SDR) score. By combining CWS-PResUNet and Demucs, our ByteMSS system ranks the 2nd on vocals score and 5th on average score in the 2021 ISMIR Music Demixing (MDX) Challenge limited training data track (leaderboard A). Our code and pre-trained models are publicly available at: https://github.com/haoheliu/2021-ISMIR-MSS-Challenge-CWS-PResUNet
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。