-
This is an issue for developing a decent framework for speaker verification under ESPnet2.
We sometimes call it 'speaker id' but just in case, on the evaluation side identification and verificatio…
-
Finalize the naming of extensions and the keywords in both the primary and extension headers.
- [x] finalize extension names
- [x] questions on the [this comment](https://github.com/NCAR/ucomp-pip…
-
how to generate the .ckpt file in x-vector
i only get final.raw by x-vector in kadli
-
Hi, when I try to run speaker_verification_plda.py all is ok in terms of the xvectors network, but when it comes to the PLDA training it fails all the time, outputting errors in relation with scipy.li…
-
**Describe the bug**
I want to summarize problems and lessons learned while using this PLDA implementation.
I observed two kinds of errors when training so far:
1. `ValueError: array must not c…
-
PLDA is often used for biometric tasks, where we need to enroll a set of models. Based on those enrollments, we can generate scores. In your setup, how the enrollment will occur?
-
Hi,
Nice job ! Thanks for sharing.
I am trying to use your code to run VoxConverse. I tried to use ground-truth VAD results for clustering, but I found that VAD results in VBx/VAD/final_syste…
-
Dear all:
I succeed to train on VoxCeleb1+2 dataset:
```python
100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████…
-
Is stage 20 used to calculate PLDA scores and the PLDA scores are in baseline/exp/result-*-*/$dset/scores?
-
New Hardware Architecture : RISC-V
This is a very experimental platform. PyAF intends to work on a not-yet-fully-manufactured hardware (#176 ).
The Sifive dev board (VisionFive 2) is planned thr…