yong xu @ seattle

你这个 s和n的长度不一样，当然不能加在一起 ---------------------------------------------------------- Yong XU From: xiaowang Date: 2018-11-07 05:01 To: yongxuUSTC/sednn CC: Subscribed Subject: [yongxuUSTC/sednn] ValueError (#20) 你好，我用的自己的语料，运行时出现一下问题 mixed_audio = s + n ValueError: operands could not be broadcast...

Hi Nick, Yes, you can use the enhanced features for ASR. But maybe you should use retraining or joint-training of your backend acoustic model for ASR. Good luck. Best regards,...

Hi Nick, Yes, there are joint SE & ASR training papers: https://www.isca-speech.org/archive/interspeech_2014/i14_0616.html https://ieeexplore.ieee.org/abstract/document/7178797/ Best regards, yong ---------------------------------------------------------- Yong XU https://sites.google.com/view/xuyong/home From: qiuqiangkong Date: 2018-07-06 03:55 To: yongxuUSTC/sednn CC: yong xu...

Hi This is according to the original noise aware training used in the robust ASR [1], which claims that "This performance can be further improved by incorporating information about the...

+1 Could you please also provide a well-trained 16kHz model? As 16kHz is the most common speech sample rate.

Have you solved it? I have the same problem as below: (myVE) yx0001@fili:~/Downloads/wavenet/wavenet$ KERAS_BACKEND=theano python wavenet.py predict with models/run_20160920_120916/config.json predict_seconds=1 Using gpu device 6: GeForce GTX TITAN X (CNMeM is...

Please use "step1_DNNenh_for16khz.m" not using "*.exe" to run the demo. High version Matlab should be OK to run the demo. Because all involved matlab functions can be updated with higher...

If you just want to get an enhanced waveform, please comment out all of related .exe files, because all of these exe files are just used to calc some metrics....

great! Yes, wav files should be in 16bit and 16khz and single channel. ---------------------------------------------------------- Yong XU https://sites.google.com/view/xuyong/home From: WallyZambotti Date: 2018-08-09 21:59 To: yongxuUSTC/DNN-Speech-enhancement-demo-tool CC: yong xu @ seattle; Comment...

The weights are here: https://drive.google.com/file/d/0B5r5bvRpQ5DRR1lIV1hpZ0RLQ0E/view On Fri, 21 Dec 2018 at 02:51, Xin Xin wrote: > hello, doc yongxu. I have a question is that I don't find > "config\se_weights50(TIMIT-16k-115NT-80H,ReLU-F6NAT-hid2500-bMFCC93-Mel40-DropoutV0.1H0.1-2out0.8S0.2N-energyN,epoch50,err46.51).mat...