arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.02865eess.AScs.LG
收藏

在基于注意力LSTM的语音可懂度等级分类系统中结合声学频谱图与调制频谱图

On combining acoustic and modulation spectrograms in an attention LSTM-based system for speech intelligibility level classification

Universidad Carlos III de Madrid(卡洛斯三世大学);Universidad Politécnica de Madrid(马德里理工大学)

Ascensión Gallardo-Antolín, Juan M. Montero

AI总结:

本文基于注意力LSTM系统,提出用逐帧调制频谱图替代紧凑特征,并探索晚期与加权池化两种融合策略,在UA-Speech上验证了互补性,其中WP融合加注意力池化效果最佳。

摘要:

语音可懂度可能受到多种因素的影响,例如嘈杂环境、信道失真或生理问题。在本工作中,我们处理的是后一种情况下语音可懂度等级的自动预测问题。基于我们之前的工作——一个为此任务设计的、基于带有注意力机制的LSTM网络的非侵入式系统,我们提出了两个主要贡献。第一个贡献是提出使用逐帧调制频谱图作为输入特征,而不是使用从它们派生出的、丢弃了重要时间信息的紧凑表示。第二个贡献是探索了将逐帧声学log-mel频谱图和调制频谱图结合到LSTM框架中的两种不同策略:决策级融合(即晚期融合)和话语级融合(即加权池化(WP)融合)。所提出的模型使用UA-Speech数据库进行评估,该数据库包含不同严重程度的构音障碍语音。一方面,结果表明,注意力LSTM网络能够充分地对调制频谱图序列进行建模,产生与log-mel频谱图情况相似的分类率。另一方面,晚期融合和WP融合这两种结合策略都优于单一特征系统,这表明逐帧log-mel频谱图和调制频谱图对于语音可懂度预测任务携带互补信息,这些信息可以被基于LSTM的架构有效利用,其中采用WP融合策略和注意力池化的系统取得了最佳结果。

英文摘要:

Speech intelligibility can be affected by multiple factors, such as noisy environments, channel distortions or physiological issues. In this work, we deal with the problem of automatic prediction of the speech intelligibility level in this latter case. Starting from our previous work, a non-intrusive system based on LSTM networks with attention mechanism designed for this task, we present two main contributions. In the first one, it is proposed the use of per-frame modulation spectrograms as input features, instead of compact representations derived from them that discard important temporal information. In the second one, two different strategies for the combination of per-frame acoustic log-mel and modulation spectrograms into the LSTM framework are explored: at decision level or late fusion and at utterance level or Weighted-Pooling (WP) fusion. The proposed models are evaluated with the UA-Speech database that contains dysarthric speech with different degrees of severity. On the one hand, results show that attentional LSTM networks are able to adequately modeling the modulation spectrograms sequences producing similar classification rates as in the case of log-mel spectrograms. On the other hand, both combination strategies, late and WP fusion, outperform the single-feature systems, suggesting that per-frame log-mel and modulation spectrograms carry complementary information for the task of speech intelligibility prediction, than can be effectively exploited by the LSTM-based architectures, being the system with the WP fusion strategy and Attention-Pooling the one that achieves best results.

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。

↑