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 2104.02606cs.CVcs.SDeess.ASeess.IV
收藏

弱监督下的视听声源检测与分离

Weakly-supervised Audio-visual Sound Source Detection and Separation

University of British Columbia(不列颠哥伦比亚大学);Vector Institute for AI(Vector人工智能研究所)

Tanzila Rahman, Leonid Sigal

摘要:

学习如何在视频的音频通道中定位和分离各个对象声音是一项艰巨的任务。目前最先进的方法从人工混合的频谱图预测音频掩码,称为混合和分离框架。我们提出了一种视听联合分割,其中网络从仅标记有对象标签的视频中学习单个对象的外观和声音。与其他最近的视觉引导音频声源分离框架不同,我们的架构可以以端到端的方式学习,并且不需要额外的监督或限界框建议。具体来说,我们在声音分离的背景下引入弱监督对象分割。我们还制定了声谱图掩模预测使用一组学习的掩模基地,结合使用对象分割的输出,一个设计,有利于分离的条件下的系数。在MUSIC数据集上的大量实验表明,我们提出的方法在视觉引导的声音声源分离和声音去噪方面优于最先进的方法。

英文摘要:

Learning how to localize and separate individual object sounds in the audio channel of the video is a difficult task. Current state-of-the-art methods predict audio masks from artificially mixed spectrograms, known as Mix-and-Separate framework. We propose an audio-visual co-segmentation, where the network learns both what individual objects look and sound like, from videos labeled with only object labels. Unlike other recent visually-guided audio source separation frameworks, our architecture can be learned in an end-to-end manner and requires no additional supervision or bounding box proposals. Specifically, we introduce weakly-supervised object segmentation in the context of sound separation. We also formulate spectrogram mask prediction using a set of learned mask bases, which combine using coefficients conditioned on the output of object segmentation , a design that facilitates separation. Extensive experiments on the MUSIC dataset show that our proposed approach outperforms state-of-the-art methods on visually guided sound source separation and sound denoising.

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。

↑