WeidiXie/VGG-Speaker-Recognition

Issues

• How to use our own dataset?

  #73 opened 2 years ago by SanaullahOfficial
  0

• Error while attempting to train the model

  #67 opened 3 years ago by aserzhqa54zwsz
  3

• How to train the model on the Voxceleb1 dataset？

  #22 opened 5 years ago by YaKaiLi
  2

• The training set contains the validation set

  #38 opened 5 years ago by tae-jun
  2

• There are some problems with voxlb2_train.txt and voxlb2_val.txt

  #48 opened 5 years ago by lyu-joe
  12

• issues in training Voxceleb1 model

  #69 opened 2 years ago by Fan0fan
  0

• Error in loading the pretrained weights

  #34 opened 5 years ago by aliakbar09a
  5

• Learning rate schedule problem

  #66 opened 3 years ago by suyuzhang
  1

• Librosa version requirements

  #59 opened 4 years ago by go2chayan
  5

• about loss and acc

  #61 opened 4 years ago by hermanseu
  6

• Preprocessing function - WAV extending

  #60 opened 4 years ago by celpas
  1

• Non-trainable params: 19,136

  #54 opened 4 years ago by llearner
  3

• TypeError: __init__() missing 2 required positional arguments: 'mode' and 'k_centers'

  #58 opened 4 years ago by preniqivjosa
  1

• about the default num_class in the function vggvox_resnet2d_icassp()

  #56 opened 4 years ago by foofybuster
  4

• Use custom feature vector instead of thin-resnet

  #55 opened 4 years ago by clintonlau
  0

• How to measure the similarity of two utterance ?

  #53 opened 4 years ago by 553566286
  1

• Validation and testing meta data

  #52 opened 4 years ago by hmen97
  3

• Got core dumped when training with GPU option

  #51 opened 4 years ago by huynguyen82
  0

• Some doubts about pre-trained model weights

  #50 opened 5 years ago by lyu-joe
  1

• TypeError: 'int' object is not callable

  #47 opened 5 years ago by IvanEvan
  0

• questions about downloading pre-trained model

  #49 opened 5 years ago by lyu-joe
  1

• ValueError: Layer #125 (named "gvlad_center_assignment"), weight <tf.Variable 'gvlad_center_assignment/kernel:0' shape=(7, 1, 512, 12) dtype=float32_ref> has shape (7, 1, 512, 12), but the saved weight has shape (10, 512, 7, 1).

  #46 opened 5 years ago by IvanEvan
  1

• how to implement a transfer learning solution for classification problems?

  #44 opened 5 years ago
  1

• Answer about hardware

  #45 opened 5 years ago
  1

• Training accuracy oscillates between 51-52% early while loss decreases slowly.

  #35 opened 5 years ago by alamnasim
  2

• using tripletLoss function

  #42 opened 5 years ago by shatealaboxiaowang
  1

• Why we need to regularize outputs during eval mode?

  #43 opened 5 years ago
  2

• any slides can share about the voxsrc workshop in interspeech 2019?

  #41 opened 5 years ago by mmxuan18
  4

• Section "Probing verification based on length" -- about verification pairs 25,020

  #40 opened 5 years ago by llearner
  2

• Training is quite slow.

  #33 opened 5 years ago by alamnasim
  8

• how to increase gpu usage?

  #39 opened 5 years ago by ArtemisZGL
  6

• two question about the vladpooling implement compare to the paper of netvlad?

  #36 opened 5 years ago by mmxuan18
  1

• problem with testing my own trained model

  #37 opened 5 years ago by aSafarpoor
  3

• ValueError: Range cannot be empty (low >= high) unless no samples are taken

  #31 opened 5 years ago by LG-SS
  1

• How good is the ghostvlad weights of the model?

  #32 opened 5 years ago by fazlekarim
  2

• error in data generation MP 2

  #29 opened 5 years ago by gogyzzz
  1

• Max pool layer strides param is different to paper

  #27 opened 5 years ago by liangyanfeng
  2

• RuntimeWarning: divide by zero encountered in true_divide

  #30 opened 5 years ago by Honghe
  3

• is there any qualitative analysis of the network learn a good embedding feature, any visualize tools for this ?

  #26 opened 5 years ago by mmxuan18
  1

• Why extend audio?

  #28 opened 5 years ago by happypanda5
  5

• Question on train loss

  #25 opened 5 years ago by seungwonpark
  4

• will you share the pretrained model which used amsoftmax, i trained myself use amsoftmax but very hard to convergence

  #23 opened 5 years ago by mmxuan18
  1

• Duration of audio file for the trained model?

  #21 opened 5 years ago by happypanda5
  1

• Different between the average and VLAD pooling

  #18 opened 5 years ago by mycrazycracy
  8

• Question on last ReLU layer for evaluating 512-dim vector

  #20 opened 5 years ago by seungwonpark
  2

• some question about the thin-ResNet?

  #19 opened 5 years ago by mmxuan18
  1

• error in data generation MP

  #17 opened 5 years ago by rohithkodali
  3

• Training under Windows

  #15 opened 5 years ago by AntonBiryukovUofC
  11

• toolkits.initialize_GPU(args) error

  #16 opened 5 years ago by WrathOfGrapes
  3

• why the wav preprocess not directly use librosa.feature.melspectrogram? what's the difference?

  #14 opened 5 years ago by mmxuan18
  5