search

yongxuUSTC / sednn

deep learning based speech enhancement using keras or pytorch, make it easy to use
http://staff.ustc.edu.cn/~jundu/The%20team/yongxu/demo/SE_DNN_taslp.html
334 stars 124 forks source link

Script for testing trained model on own sample recording #16

Closed akshayaCap closed 6 years ago

akshayaCap commented 6 years ago

Hi Yong Xu, I have finished training model. I am stuck on testing the output of trained model on new recording. Is there a script for this? Or if someone could point out the changes in prepare_data.py and main_dnn.py for this. The issue is I have a recording with noise so cannot define signal to noise ratio.

Thank you Akshaya

qiuqiangkong commented 6 years ago

Hi Akshaya,

For testing the new audio, you may extract the features first, then apply the model on the feature. Please see the code here for inference https://github.com/yongxuUSTC/sednn/blob/master/mixture2clean_dnn/main_dnn.py#L172

Best wishes,

Qiuqiang

From: akshayaCap notifications@github.com Sent: 31 July 2018 07:06:25 To: yongxuUSTC/sednn Cc: Subscribed Subject: [yongxuUSTC/sednn] Script for testing trained model on own sample recording (#16)

Hi Yong Xu, I have finished training model. I am stuck on testing the output of trained model on new recording. Is there a script for this? Or if someone could point out the changes in prepare_data.py and main_dnn.py for this. The issue is I have a recording with noise so cannot define signal to noise ratio.

Thank you Akshaya

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/yongxuUSTC/sednn/issues/16, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AMt5yc4vXuoOO8rNMruGoGiG_5oneoCOks5uL_PhgaJpZM4Vnpl3.