pirxus / personalvad Goto Github PK

Hi, @pirxus really kind of shared implementation,
It would be great to user if you could share the checkpoint for fast evaluation.

Hi, thank you for uploading your code!
I want to use the pretrained models to test my own data, but it seems that there is something wrong during the process of feature extraction. So I'd like to ask how should I use the model to test my own data?

Hi,

I am currently trying to replicate the results from the Personal VAD paper, and I am having some issues with the model not properly distinguishing between target speaker and non-target speaker speech, with a heavy bias towards target speaker speech.
Have you had any such issues?

Best regards,
Holger Severin Bovbjerg

Hi, will you try to implement the personalVAD 2.0 ?
Personal VAD 2.0: Optimizing Personal Voice Activity Detection for On-Device Speech Recognition

Hi, thank you for your personal VAD code!

when i try to extract feature (command: bash prepare_dataset_features.sh 3), i encounter some problem as below:

File "/usr/lib/python3.6/multiprocessing/pool.py", line 119, in worker
result = (True, func(*args, **kwds))
File "/usr/lib/python3.6/multiprocessing/pool.py", line 47, in starmapstar
return list(itertools.starmap(args[0], args[1]))
File "src/extract_features.py", line 176, in extract_features
for utt_id, (sr, arr) in wav_scp:
File "/home/oscar/.local/lib/python3.6/site-packages/kaldiio/highlevel.py", line 136, in iter
k, v = next(self.generator)
File "/home/oscar/.local/lib/python3.6/site-packages/kaldiio/matio.py", line 127, in load_scp_sequential
mat = _load_mat(arkfd, offset, slices, endian=endian)
File "/home/oscar/.local/lib/python3.6/site-packages/kaldiio/matio.py", line 331, in _load_mat
array = read_kaldi(fd, endian)
File "/home/oscar/.local/lib/python3.6/site-packages/kaldiio/matio.py", line 441, in read_kaldi
array = read_ascii_mat(fd)
File "/home/oscar/.local/lib/python3.6/site-packages/kaldiio/matio.py", line 605, in read_ascii_mat
assert len(string) != 0
AssertionError

Thank you for your help. I look forward to hearing from you.

Sincerely,