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 2005.11185cs.CLcs.SDeess.AS
收藏

基于部分假设选择的低延迟序列到序列语音识别和翻译

Low-Latency Sequence-to-Sequence Speech Recognition and Translation by Partial Hypothesis Selection

Maastricht University(马斯特里赫特大学)

Danni Liu, Gerasimos Spanakis, Jan Niehues

摘要:

编码器-解码器模型为诸如语音识别和翻译之类的序列到序列任务提供通用架构。虽然离线系统通常根据字出错率(WER)和BLEU等质量指标进行评估,但在许多实际用例中,延迟也是一个关键因素。我们提出了三个延迟减少技术基于块的增量推理和评估其效率的准确性延迟权衡。在300小时的How2数据集上,我们通过牺牲1%的WER(6%的可靠性)将延迟减少了83%,达到0.8秒。与离线转录相比。虽然我们的实验使用了Transformer,但假设选择策略适用于其他编码器-解码器模型。为了避免昂贵的重新计算,我们使用单向参与编码器。在对部分序列的适应过程之后,单向模型与原始模型表现一致。我们进一步表明,我们的方法也适用于低延迟语音翻译。在How2英语-葡萄牙语语音翻译上,我们将延迟减少到0.7秒(-84% rel)。同时损失了2.4个BLEU点(5%相对值)。与离线系统相比。

英文摘要:

Encoder-decoder models provide a generic architecture for sequence-to-sequence tasks such as speech recognition and translation. While offline systems are often evaluated on quality metrics like word error rates (WER) and BLEU, latency is also a crucial factor in many practical use-cases. We propose three latency reduction techniques for chunk-based incremental inference and evaluate their efficiency in terms of accuracy-latency trade-off. On the 300-hour How2 dataset, we reduce latency by 83% to 0.8 second by sacrificing 1% WER (6% rel.) compared to offline transcription. Although our experiments use the Transformer, the hypothesis selection strategies are applicable to other encoder-decoder models. To avoid expensive re-computation, we use a unidirectionally-attending encoder. After an adaptation procedure to partial sequences, the unidirectional model performs on-par with the original model. We further show that our approach is also applicable to low-latency speech translation. On How2 English-Portuguese speech translation, we reduce latency to 0.7 second (-84% rel.) while incurring a loss of 2.4 BLEU points (5% rel.) compared to the offline system.

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。

↑