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 2102.01326eess.AScs.LGcs.SD
收藏

多模态注意力融合的目标说话人提取

Multimodal Attention Fusion for Target Speaker Extraction

NTT Corporation(日本电报电话公司)

Hiroshi Sato, Tsubasa Ochiai, Keisuke Kinoshita, Marc Delcroix, Tomohiro Nakatani, Shoko Araki

摘要:

目标说话人提取是利用声音、视觉或位置信息从混合语音中提取出目标说话人的语音,受到了广泛的关注。最近提出了一种视听目标说话人提取方法,该方法利用互补的视听线索提取目标语音。虽然视听目标说话人提取提供了一个更稳定的性能比单一模态的方法模拟数据,其适应现实情况还没有得到充分的探讨,以及对真实记录的混合物的评价。处理现实情况的主要问题之一是如何使系统对线索损坏具有鲁棒性,因为在真实记录中,两个线索可能不是同样可靠的,例如,视觉线索可能受到遮挡的影响。在这项工作中,我们提出了一种新的多模态融合的注意力机制及其训练方法,能够有效地捕获线索的可靠性并对更可靠的线索进行加权。我们的建议提高信号失真比(SDR)1.0 dB的模拟数据比传统的融合机制。此外,我们还记录了一个视听数据集的同时语音与现实的视觉线索腐败,并表明视听目标说话人提取与我们的建议成功地工作在真实的数据。

英文摘要:

Target speaker extraction, which aims at extracting a target speaker's voice from a mixture of voices using audio, visual or locational clues, has received much interest. Recently an audio-visual target speaker extraction has been proposed that extracts target speech by using complementary audio and visual clues. Although audio-visual target speaker extraction offers a more stable performance than single modality methods for simulated data, its adaptation towards realistic situations has not been fully explored as well as evaluations on real recorded mixtures. One of the major issues to handle realistic situations is how to make the system robust to clue corruption because in real recordings both clues may not be equally reliable, e.g. visual clues may be affected by occlusions. In this work, we propose a novel attention mechanism for multi-modal fusion and its training methods that enable to effectively capture the reliability of the clues and weight the more reliable ones. Our proposals improve signal to distortion ratio (SDR) by 1.0 dB over conventional fusion mechanisms on simulated data. Moreover, we also record an audio-visual dataset of simultaneous speech with realistic visual clue corruption and show that audio-visual target speaker extraction with our proposals successfully work on real data.

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。

↑