-
Hi,
I dont see any data augmentation being done here,
what kind of data augmentation would be suitable for BonaFide/Spoof data?
Regards,
Leonard
-
微博内容精选
-
为什么使用oc-softmax之后得分,正样本的得分是再-1左右,而负样本是在1左右。但是使用softmax,得分确是正样本在1左右,负样本在0左右?
-
Hi!
I have a few questions, didn't find out answers in the repo, so let me ask them here:
Are models resistant to voice spoofing attacks (replay, voice cloning/conversion/ tts) for speaker verificat…
-
Hello, thanks for sharing. I've read your one-class classification paper based on AM-Softmax and that was a very nice approach.
I would like to ask a general question regarding your repositories. I…
-
Sir, I have read your article.And i have some questions.i can't understand this Figure 4.How did you get this figure4.The test set has no tags,so how did you know which spoofing attack method was used…
-
**Is your feature request related to a problem? Please describe.**
It's not related to a problem. We are using librosa as our go-to tool for audio feature extraction. We love that we get an API to …
-
applying irspace
space IR ({'filter': 'NiceDrumRoom-16000.ir', 'wet': 100})
only 1 file matching .*NiceDrumRoom-16000.ir.*
using /root/data/acoustic-simulator-master/impulse-responses/spaces…
-
As in the model.py file like: "/project/03-asvspoof-mega/lfcc-lcnn-lstmsum-p2s/01/model.py ", I don't understand why use def normalize_target(self, y): in line #272, I thought the target data was the…
-
def _parse_line(self, line):
tokens = line.strip().split(' ')
if self.is_eval:
return ASVFile(speaker_id=tokens[0],
file_name=tokens[1],
…