-
Hello:
Hello, when I run the AMI dataset program in speaker dialogue in speech train, I consider whether I can use speech train to train an embeddings about d-vector
So I want to ask if speechtrain …
-
Hello,
What are the differences between CNN custom implementation (Conv1d, BatchNorm1d...) and PyTorch implementation (torch.nn.Conv1d, torch.nn.BatchNorm1d...). I've noticed that custom cnn are …
-
hi
The ouput of `python train_speaker_embeddings.py hyperparams/train_x_vectors.yaml` doesnot have the file of "mean_var_norm_emb.ckpt", which will be used in "verification_plda_xvector.yaml"; is it(…
-
I'm trying to follow the recipe for speaker diarization on the AMI dataset (https://github.com/speechbrain/speechbrain/tree/develop/recipes/AMI/Diarization) but unfortunately without success. Here's t…
-
Dear All,
I tried to run "python train_speaker_embeddings.py hparams/train_ecapa_tdnn_big.yaml" after finish the training process. But I can't find the "mean_var_norm_emb.ckpt" file. How to generat…
-
Well, after the issue with the installation (now closed) I have another problem.
I am trying to create a model that distinguishes between only a few languages (instead of the 20+ languages the Comm…
-
Hi, Thanks for great job. I have tried to use Mixup Learning Strategies,
but I got an Error which says:
Mixup object has no attribute 'lam'
I think this may be an implementation bug.
-
Hi,
Working on data augmentation experiments for speaker verification and would like to test my ideas on speechbrain.
Currently recipes/VoxCeleb/SpeakerRec/hparams contains:
train_ecapa_tdnn.yaml…
-
During [evaluation](https://github.com/TaoRuijie/ECAPATDNN/blob/main/ECAPAModel.py#L82), the current implementation calculates the similarity scores one by one using a for loop, that could be slow whe…
-
Hello! I tried to retrain your model and it's got 0.95% EER with score normalization instead of 0.69% EER. The training was done with 4 TITAN X using data-parallel backend on vox1+vox2, batch size was…