已有语音转文字服务:
(尚无中文)Zoom自带:
https://support.zoom.us/hc/zh-cn/articles/115004794983-自动转换云录制
谷歌云:
https://cloud.google.com/speech-to-text?hl=zh-cn
腾讯AI:
https://ai.qq.com/product/aaiasr.shtml
讯飞:
https://www.xfyun.cn/services/voicedictation
依图:
https://speech.yitutech.com/
百度:
https://ai.baidu.com/tech/speech
搜狗:
http://dictation.sogou.com/#/
IBM Watson:
https://github.com/watson-developer-cloud/speech-to-text-websockets-python
Microsoft:
https://azure.microsoft.com/zh-cn/services/cognitive-services/speech-to-text/
Amazon:
https://aws.amazon.com/cn/transcribe/
知乎回答:
https://www.zhihu.com/question/20124290
,
https://www.zhihu.com/question/21552953
待补充...
开源模型:
https://github.com/seracc/masr
基于PyTorch,有tensorboardX可视化
预训练模型:
https://github.com/seracc/masr/releases/download/m1/gated-conv.pth
基建搭的不错,文档比较全:
https://github.com/seracc/masr/tree/master/docs
, 4天前刚上传比较新,训练数据集只有一个AISHELL-1
https://github.com/audier/DeepSpeechRecognition
基于Tensorflow+Keras
预训练模型: 没找到
对应博文:
https://blog.csdn.net/chinatelecom08/article/details/82557715
文档还算全,最后更新2019/03,数据集: thchs-30, AISHELL, primewords, st-cmd
待补充...
数据集:
http://www.openslr.org/resources.php
搜Mandarin关键词 貌似挺全了
http://www.aishelltech.com/aishell_2
AISHELL_2v貌似比较新 需要申请使用