Accent-Recognition

时间:2021-04-18 17:37:04
【文件属性】:
文件名称:Accent-Recognition
文件大小:10.65MB
文件格式:ZIP
更新时间:2021-04-18 17:37:04
Python 口音识别 重读英语语音识别的数据准备脚本和培训管道。 取决于环境 Kaldi(与数据准备相关的功能脚本) Espnet Google SentencePiece(pip3安装语句) 在path.sh文件中修改espnet的安装地址 使用说明 资料准备 All the data used in the experiment are stored in the `data` directory, in which train is used for training, valid is the verification set, cv_all and test are used for testing respectively. In order to better reproduce my experimental results, you can download the data
【文件预览】:
Accent-Recognition-main
----models()
--------e2e_asr_transformer_only_accent_with_attention.py(12KB)
--------e2e_asr_transformer_only_accent.py(12KB)
----run_only_accent.sh(9KB)
----conf()
--------e2e_asr_transformer_only_accent.yaml(925B)
--------espnet_lm.yaml(292B)
--------e2e_asr_transformer_only_accent_with_attention.yaml.yaml(939B)
--------espnet_specaug.yaml(322B)
--------espnet_decode.yaml(122B)
--------espnet_train.yaml(1KB)
----run_accent160_asr.sh(10KB)
----data()
--------cv_all()
--------lang()
--------train()
--------test()
--------valid()
----local()
--------tools()
--------files()
----path.sh(785B)
----README.md(4KB)

网友评论