arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2109.04574cs.CL
收藏

Speechformer:减少直接语音翻译中的信息损失

Speechformer: Reducing Information Loss in Direct Speech Translation

Fondazione Bruno Kessler(布鲁诺·凯斯勒基金会);University of Trento(特伦托大学)

Sara Papi, Marco Gaido, Matteo Negri, Marco Turchi

AI总结:

针对直接语音翻译中因Transformer处理长音频序列需进行初始有损压缩导致信息丢失的问题,提出Speechformer架构,通过减少注意力层内存使用避免初始压缩,在更高层按语言学标准聚合信息,在标准语料库和低资源场景下分别提升0.8和4.0 BLEU。

摘要:

基于Transformer的模型在包括语音翻译在内的许多研究领域中取得了最先进的性能,其受欢迎程度日益增加。然而,Transformer相对于输入序列长度的二次复杂度阻碍了其直接应用于通常由长序列表示的音频信号。目前的解决方案诉诸于基于原始音频特征的固定采样的初始次优压缩。因此,架构中的高层无法获取潜在有用的语言信息。为了解决这个问题,我们提出了Speechformer,这是一种架构,由于减少了注意力层中的内存使用,它避免了初始有损压缩,并且仅根据更充分的语言学标准在更高层次上聚合信息。在三个语言对(en->de/es/nl)上的实验表明了我们解决方案的有效性,在标准MuST-C语料库上获得了高达0.8 BLEU的增益,在低资源场景下获得了高达4.0 BLEU的增益。

英文摘要:

Transformer-based models have gained increasing popularity achieving state-of-the-art performance in many research fields including speech translation. However, Transformer's quadratic complexity with respect to the input sequence length prevents its adoption as is with audio signals, which are typically represented by long sequences. Current solutions resort to an initial sub-optimal compression based on a fixed sampling of raw audio features. Therefore, potentially useful linguistic information is not accessible to higher-level layers in the architecture. To solve this issue, we propose Speechformer, an architecture that, thanks to reduced memory usage in the attention layers, avoids the initial lossy compression and aggregates information only at a higher level according to more informed linguistic criteria. Experiments on three language pairs (en->de/es/nl) show the efficacy of our solution, with gains of up to 0.8 BLEU on the standard MuST-C corpus and of up to 4.0 BLEU in a low resource scenario.

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。

↑