首页> 外文会议>2017 IEEE Automatic Speech Recognition and Understanding Workshop >JHU Kaldi system for Arabic MGB-3 ASR challenge using diarization, audio-transcript alignment and transfer learning
【24h】

JHU Kaldi system for Arabic MGB-3 ASR challenge using diarization, audio-transcript alignment and transfer learning

机译:JHU Kaldi系统,用于阿拉伯语MGB-3 ASR挑战,采用了对白,音频记录对齐和转移学习

获取原文
获取原文并翻译 | 示例

摘要

This paper describes the JHU team's Kaldi system submission to the Arabic MGB-3: The Arabic speech recognition in the Wild Challenge for ASRU-2017. We use a weights transfer approach to adapt a neural network trained on the out-of-domain MGB-2 multi-dialect Arabic TV broadcast corpus to the MGB-3 Egyptian YouTube video corpus. The neural network has a TDNN-LSTM architecture and is trained using lattice-free maximum mutual information (LF-MMI) objective followed by sMBR discriminative training. For supervision, we fuse transcripts from 4 independent transcribers into confusion network training graphs. We also describe our own approach for speaker diarization and audio-transcript alignment. We use this to prepare lightly supervised transcriptions for training the seed system used for adaptation to MGB-3. Our primary submission to the challenge gives a multi-reference WER of 32.78% on the MGB-3 test set.
机译:本文描述了JHU团队的Kaldi系统提交给阿拉伯语MGB-3:ASRU-2017狂野挑战赛中的阿拉伯语语音识别。我们使用权重转移方法,将在域外MGB-2多方言阿拉伯电视广播语料库上训练的神经网络改编为MGB-3埃及YouTube视频语料库。该神经网络具有TDNN-LSTM架构,并使用无格最大互信息(LF-MMI)目标进行训练,然后进行sMBR判别训练。为了进行监督,我们将来自4个独立抄写员的成绩单融合到混淆网络训练图中。我们还描述了自己的说话人区分和音频记录对齐方式。我们用它来准备轻微监督的转录,以训练用于适应MGB-3的种子系统。我们对挑战的主要提交意见在MGB-3测试仪上给出了32.78%的多参考WER。

著录项

相似文献

  • 外文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号