WeidiXie / VGG-Speaker-Recognition

Utterance-level Aggregation For Speaker Recognition In The Wild

WeidiXie/VGG-Speaker-Recognition Issues

How to use our own dataset?
Updated 2 years ago
Error while attempting to train the model
Updated 2 years ago3
How to train the model on the Voxceleb1 dataset？
Closed 5 years ago2
The training set contains the validation set
Closed 5 years ago2
There are some problems with voxlb2_train.txt and voxlb2_val.txt
Closed 4 years ago12
issues in training Voxceleb1 model
Updated 2 years ago
Error in loading the pretrained weights
Closed 5 years ago5
Learning rate schedule problem
Closed 3 years ago1
Librosa version requirements
Closed 4 years ago5
about loss and acc
Closed 4 years ago6
Preprocessing function - WAV extending
Closed 4 years ago1
Non-trainable params: 19,136
Closed 4 years ago3
TypeError: __init__() missing 2 required positional arguments: 'mode' and 'k_centers'
Updated 4 years ago1
about the default num_class in the function vggvox_resnet2d_icassp()
Closed 4 years ago4
Use custom feature vector instead of thin-resnet
Updated 4 years ago
How to measure the similarity of two utterance ?
Closed 4 years ago1
Validation and testing meta data
Closed 4 years ago3
Got core dumped when training with GPU option
Closed 4 years ago
Some doubts about pre-trained model weights
Closed 4 years ago1
TypeError: 'int' object is not callable
Closed 4 years ago
questions about downloading pre-trained model
Closed 4 years ago1
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).
Closed 4 years ago1
how to implement a transfer learning solution for classification problems?
Closed 4 years ago1
Answer about hardware
Closed 4 years ago1
Training accuracy oscillates between 51-52% early while loss decreases slowly.
Closed 4 years ago2
using tripletLoss function
Closed 4 years ago1
Why we need to regularize outputs during eval mode?
Closed 4 years ago2
any slides can share about the voxsrc workshop in interspeech 2019?
Closed 5 years ago4
Section "Probing verification based on length" -- about verification pairs 25,020
Closed 5 years ago2
Training is quite slow.
Closed 5 years ago8
how to increase gpu usage?
Closed 5 years ago6
two question about the vladpooling implement compare to the paper of netvlad?
Closed 5 years ago1
problem with testing my own trained model
Closed 5 years ago3
ValueError: Range cannot be empty (low >= high) unless no samples are taken
Closed 5 years ago1
How good is the ghostvlad weights of the model?
Closed 5 years ago2
error in data generation MP 2
Closed 5 years ago1
Max pool layer strides param is different to paper
Closed 5 years ago2
RuntimeWarning: divide by zero encountered in true_divide
Closed 5 years ago3
is there any qualitative analysis of the network learn a good embedding feature, any visualize tools for this ?
Closed 5 years ago1
Why extend audio?
Closed 5 years ago5
Question on train loss
Closed 5 years ago4
will you share the pretrained model which used amsoftmax, i trained myself use amsoftmax but very hard to convergence
Closed 5 years ago1
Duration of audio file for the trained model?
Closed 5 years ago1
Different between the average and VLAD pooling
Closed 5 years ago8
Question on last ReLU layer for evaluating 512-dim vector
Closed 5 years ago2
some question about the thin-ResNet?
Closed 5 years ago1
error in data generation MP
Closed 5 years ago3
Training under Windows
Closed 5 years ago11
toolkits.initialize_GPU(args) error
Closed 5 years ago3
why the wav preprocess not directly use librosa.feature.melspectrogram? what's the difference?
Closed 5 years ago5