breizhn/DTLN-aec

Issues

• The model based on 'DTLN' project and your paper about 'DTLN-aec' can't be trained as good as your pretrained one.

  #27 opened 2 years ago by steven8274
  5

• Is this a mistake in the paper? or Am i misunderstood it ?

  #5 opened 4 years ago by fhahaha
  2

• How to training DTLN-aec model

  #26 opened 2 years ago by undermore
  0

• Some questions about concatenate operation in DTLN-aec model?

  #25 opened 2 years ago by xk2016
  0

• Problem in duplicating the training process

  #4 opened 4 years ago by angusxu458
  11

• Does DTLN-aec also contain the noise suppression?

  #24 opened 2 years ago by cloudvc
  0

• Can you open source your crowdsourced test data and results?

  #19 opened 3 years ago by zuowanbushiwo
  0

• Just Questions this time :)

  #23 opened 3 years ago by StuartIanNaylor
  0

• about repository construction plan

  #22 opened 3 years ago by jet-yangqs
  0

• H

  #21 opened 3 years ago by FragrantRookie
  0

• can it be used in real time voice?

  #9 opened 4 years ago by dttlgotv
  4

• About the training target: nearend speech

  #20 opened 3 years ago by liziru
  0

• Could DTLN-aec process audio frame real-time on mobile?

  #3 opened 4 years ago by DyncEric
  2

• How to use the model to generate the echo cancelled file.

  #18 opened 3 years ago by victkid
  0

• Looking forward to more updates about training details.

  #17 opened 3 years ago by liziru
  0

• the SNR loss is inf or nan in the training stage

  #15 opened 4 years ago by Xiaowei-coder
  0

• up-sampling the phone data to 16k and then down-sampling to 8k Model effect will be better

  #16 opened 3 years ago by JuneRen
  2

• Training problems in double talk situations

  #13 opened 4 years ago by LXP-Never
  1

• Can I get some files for quantization of the model?

  #14 opened 4 years ago by takhyun12
  0

• Keras models needed

  #12 opened 4 years ago by abu100
  0

• speech is totolly removed.

  #1 opened 4 years ago by dttlgotv
  9

• Making Echo Cancellation working on real time voice

  #11 opened 4 years ago by jyoonab
  0

• MODEL FOR aec

  #10 opened 4 years ago by Anthony2018
  0

• The influence of the hyperparameter encoder_size

  #6 opened 4 years ago by garrywrj
  1

• Failed on mobile device

  #8 opened 4 years ago by DyncEric
  2

• Loss decrease for a few and sudden explodes to inf

  #7 opened 4 years ago by angusxu458
  2

• where aru your training data for this project?

  #2 opened 4 years ago by dttlgotv
  0