SpeechStew:简单地混合所有可用的语音识别数据来训练一个大型神经网络
SpeechStew: Simply Mix All Available Speech Recognition Data to Train One Large Neural Network
摘要:
我们提出了SpeechStew,一个语音识别模型,该模型是在各种公开的语音识别数据集的组合上训练的:AMI,广播新闻,共同的声音,LibriSpeech,Switchboard/Fisher,Tedews和华尔街日报。SpeechStew只是将所有这些数据集混合在一起,而没有对数据集进行任何特殊的重新加权或重新平衡。SpeechStew在各种任务中实现SoTA或接近SoTA的结果,而无需使用外部语言模型。我们的结果包括AMI-IHM上的9.0\% WER,Switchboard上的4.7\% WER,CallHome上的8.3\% WER,以及WSJ上的1.3\%,这显著优于具有强外部语言模型的先前工作。我们还证明了SpeechStew学习强大的迁移学习表示。我们在嘈杂的低资源语音数据集CHiME-6上微调SpeechStew。在没有语言模型的情况下,我们实现了38.9\%的WER,相比之下,在有语言模型的情况下,强HMM基线的WER为38.6\%。
英文摘要:
We present SpeechStew, a speech recognition model that is trained on a combination of various publicly available speech recognition datasets: AMI, Broadcast News, Common Voice, LibriSpeech, Switchboard/Fisher, Tedlium, and Wall Street Journal. SpeechStew simply mixes all of these datasets together, without any special re-weighting or re-balancing of the datasets. SpeechStew achieves SoTA or near SoTA results across a variety of tasks, without the use of an external language model. Our results include 9.0\% WER on AMI-IHM, 4.7\% WER on Switchboard, 8.3\% WER on CallHome, and 1.3\% on WSJ, which significantly outperforms prior work with strong external language models. We also demonstrate that SpeechStew learns powerful transfer learning representations. We fine-tune SpeechStew on a noisy low resource speech dataset, CHiME-6. We achieve 38.9\% WER without a language model, which compares to 38.6\% WER to a strong HMM baseline with a language model.
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。