Almost Unsupervised Text to Speech and Automatic Speech Recognition
Abstract: 无监督方法，只需要利用几百对文本—语音对和额外的无标签的数据，提供给TTS和ASR components: 1.a denosising auto-encoder 2. 双机制训练；TTS是把text y转成语音x，ASR把利用x和y进行训练，反之亦然 3. 双向序列建模，主要解决长语音序列和文本序列在训练过程中出现的错误传播问题 4.一个unified model 包含 TT
相关文章
- 1. Voice Conversion by Cascading Automatic Speech Recognition and Text-to-Speech Synthesis with Prosody
- 2. 5 Open Source Speech Recognition/Speech-to-Text Systems
- 3. Visual Speech Recognition: Lip Segmentation and Mapping
- 4. Deep Speech 2: End-to-End Speech Recognition in English and Mandarin
- 5. SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition
- 6. Azure Cognitive Services- Speech To Text
- 7. Deep Speech 2 : End-to-End Speech Recognition in English and Mandarin论文笔记
- 8. Food Log with Speech Recognition and NLP
- 9. FastSpeech: Fast, Robust and Controllable Text to Speech
- 10. （IS 19）wav2vec: Unsupervised Pre-training for Speech Recognition
- 更多相关文章...
- • SVG
- SVG 教程 - • XSL-FO table-and-caption 对象 - XSL-FO 教程
- • RxJava操作符（七）Conditional and Boolean
- • 为了进字节跳动，我精选了29道Java经典算法题，带详细讲解
相关标签/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
欢迎关注本站公众号,获取更多信息
相关文章
- 1. Voice Conversion by Cascading Automatic Speech Recognition and Text-to-Speech Synthesis with Prosody
- 2. 5 Open Source Speech Recognition/Speech-to-Text Systems
- 3. Visual Speech Recognition: Lip Segmentation and Mapping
- 4. Deep Speech 2: End-to-End Speech Recognition in English and Mandarin
- 5. SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition
- 6. Azure Cognitive Services- Speech To Text
- 7. Deep Speech 2 : End-to-End Speech Recognition in English and Mandarin论文笔记
- 8. Food Log with Speech Recognition and NLP
- 9. FastSpeech: Fast, Robust and Controllable Text to Speech
- 10. （IS 19）wav2vec: Unsupervised Pre-training for Speech Recognition